mirror/mariadb
1
0
Fork 0
mirror of https://github.com/MariaDB/server.git synced 2025-01-17 12:32:27 +01:00
Code Issues Projects Releases Packages Wiki Activity Actions
mariadb/storage/innobase/buf/buf0flu.cc

2549 lines
77 KiB
C++
Raw Normal View History

initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
/*****************************************************************************
5.6.36
2017-05-15 16:17:16 +02:00
Copyright (c) 1995, 2017, Oracle and/or its affiliates. All Rights Reserved.
MDEV-27499 Performance regression in log_checkpoint_margin() In commit 4c3ad24413f234599eda27f4958dd3ff21df3203 (MDEV-27416) an unnecessarily strict wait condition was introduced in the function buf_flush_wait(). Most callers actually only care that the pages have been flushed, not that a checkpoint has completed. Only in the buf_flush_sync() call for log resizing, we might care about the log checkpoint. But, in fact, srv_prepare_to_delete_redo_log_file() is explicitly disabling checkpoints. So, we can simply remove the unnecessary wait loop. Thanks to Krunal Bauskar for reporting this performance regression that we failed to repeat in our testing.
2022-01-17 15:56:07 +01:00
Copyright (c) 2013, 2022, MariaDB Corporation.
Merge branch '10.0' into 10.1 referenced_by_foreign_key2(), needed for InnoDB to compile, was taken from 10.0-galera
2015-09-03 12:58:41 +02:00
Copyright (c) 2013, 2014, Fusion-io
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
This program is free software; you can redistribute it and/or modify it under
the terms of the GNU General Public License as published by the Free Software
Foundation; version 2 of the License.
This program is distributed in the hope that it will be useful, but WITHOUT
ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS
FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with
this program; if not, write to the Free Software Foundation, Inc.,
Update FSF address
2019-05-11 18:25:02 +02:00
51 Franklin Street, Fifth Floor, Boston, MA 02110-1335 USA
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
*****************************************************************************/
/**************************************************//**
@file buf/buf0flu.cc
The database buffer buf_pool flush algorithm
Created 11/11/1995 Heikki Tuuri
*******************************************************/
Remove many redundant #include from InnoDB
2018-11-19 10:42:14 +01:00
#include "univ.i"
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
#include <my_service_manager.h>
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
#include <mysql/service_thd_wait.h>
MDEV-13670/MDEV-14550 Error log flood : "InnoDB: page_cleaner: 1000ms intended loop took N ms. The settings might not be optimal." Silence the error log output that was introduced in MySQL 5.7 (MariaDB 10.2.2) if log_warnings=2 or less. We should still figure out what these messages really indicate and how to solve the problems. pc_sleep_if_needed(): Add a parameter for the current time, so that there will be fewer successive calls to ut_time_ms() with no I/O between them. buf_flush_page_cleaner_coordinator(): Exit the first loop whenever shutdown has been requested. At the start of the loop, call ut_time_ms() only once. Do not display the message if log_warnings=2 or less.
2017-12-05 11:50:35 +01:00
#include <sql_class.h>
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
#include "buf0flu.h"
#include "buf0buf.h"
#include "buf0checksum.h"
MDEV-12353: Write log by mtr_t member functions only mtr_t::log_write_low(): Replaces mlog_write_initial_log_record_low(). mtr_t::log_file_op(): Replaces fil_op_write_log(). mtr_t::free(): Write MLOG_INIT_FREE_PAGE. mtr_t::init(): Write MLOG_INIT_FILE_PAGE2. mtr_t::page_create(): Write record about the partial initialization of an index page. mlog_catenate_ulint(), mlog_catenate_string(), mlog_open(), mlog_close(): Remove.
2020-01-21 14:46:20 +01:00
#include "buf0dblwr.h"
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
#include "srv0start.h"
#include "page0zip.h"
#include "fil0fil.h"
MDEV-21132 Remove buf_page_t::newest_modification At each mini-transaction commit, the log sequence number of the mini-transaction must be written to each modified page, so that it will be available in the FIL_PAGE_LSN field when the page is being read in crash recovery. InnoDB was unnecessarily allocating redundant storage for the field, in buf_page_t::newest_modification. Let us access FIL_PAGE_LSN directly. Furthermore, on ALTER TABLE...IMPORT TABLESPACE, let us write 0 to FIL_PAGE_LSN instead of using log_sys.lsn. buf_flush_init_for_writing(), buf_flush_update_zip_checksum(), fil_encrypt_buf_for_full_crc32(), fil_encrypt_buf(), fil_space_encrypt(): Remove the parameter lsn. buf_page_get_newest_modification(): Merge with the only caller. buf_tmp_reserve_compression_buf(), buf_tmp_page_encrypt(), buf_page_encrypt(): Define static in the same compilation unit with the only caller. PageConverter::m_current_lsn: Remove. Write 0 to FIL_PAGE_LSN on ALTER TABLE...IMPORT TABLESPACE.
2019-11-25 08:39:51 +01:00
#include "log0crypt.h"
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
#include "srv0mon.h"
Enhancement: Change atomic_writes table option to enum type. Now every file can either use atomic writes, not use it or use default. SYNTAX: ATOMIC_WRITES=['DEFAULT','ON','OFF'] Idea here is to be able to define innodb_doublewrite = 1 but with following rules: ATOMIC_WRITES='DEFAULT' - if innodb_use_atomic_writes = 1, we do not write to doublewrite buffer the changes if innodb_use_atomic_writes = 0, we write to doublewrite buffer ATOMIC_WRITES='ON' - do not write to doublewrite buffer ATOMIC_WRITES='OFF' - write to doublewrite buffer Note that doublewrite buffer can't be used if innodb_doublewrite = 0.
2014-01-10 11:11:36 +01:00
#include "fil0pagecompress.h"
MDEV-21132 Remove buf_page_t::newest_modification At each mini-transaction commit, the log sequence number of the mini-transaction must be written to each modified page, so that it will be available in the FIL_PAGE_LSN field when the page is being read in crash recovery. InnoDB was unnecessarily allocating redundant storage for the field, in buf_page_t::newest_modification. Let us access FIL_PAGE_LSN directly. Furthermore, on ALTER TABLE...IMPORT TABLESPACE, let us write 0 to FIL_PAGE_LSN instead of using log_sys.lsn. buf_flush_init_for_writing(), buf_flush_update_zip_checksum(), fil_encrypt_buf_for_full_crc32(), fil_encrypt_buf(), fil_space_encrypt(): Remove the parameter lsn. buf_page_get_newest_modification(): Merge with the only caller. buf_tmp_reserve_compression_buf(), buf_tmp_page_encrypt(), buf_page_encrypt(): Define static in the same compilation unit with the only caller. PageConverter::m_current_lsn: Remove. Write 0 to FIL_PAGE_LSN on ALTER TABLE...IMPORT TABLESPACE.
2019-11-25 08:39:51 +01:00
#ifdef HAVE_LZO
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
# include "lzo/lzo1x.h"
#elif defined HAVE_SNAPPY
# include "snappy-c.h"
MDEV-21132 Remove buf_page_t::newest_modification At each mini-transaction commit, the log sequence number of the mini-transaction must be written to each modified page, so that it will be available in the FIL_PAGE_LSN field when the page is being read in crash recovery. InnoDB was unnecessarily allocating redundant storage for the field, in buf_page_t::newest_modification. Let us access FIL_PAGE_LSN directly. Furthermore, on ALTER TABLE...IMPORT TABLESPACE, let us write 0 to FIL_PAGE_LSN instead of using log_sys.lsn. buf_flush_init_for_writing(), buf_flush_update_zip_checksum(), fil_encrypt_buf_for_full_crc32(), fil_encrypt_buf(), fil_space_encrypt(): Remove the parameter lsn. buf_page_get_newest_modification(): Merge with the only caller. buf_tmp_reserve_compression_buf(), buf_tmp_page_encrypt(), buf_page_encrypt(): Define static in the same compilation unit with the only caller. PageConverter::m_current_lsn: Remove. Write 0 to FIL_PAGE_LSN on ALTER TABLE...IMPORT TABLESPACE.
2019-11-25 08:39:51 +01:00
#endif
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
/** Number of pages flushed via LRU. Protected by buf_pool.mutex.
Also included in buf_flush_page_count. */
MDEV-18582: Extend SHOW STATUS LIKE 'Innodb_%' Many InnoDB internal variables and counters were only exposed in an unstructured fashion via SHOW ENGINE INNODB STATUS. Expose more variables via SHOW STATUS. Many of these were exported in XtraDB. Also, introduce SHOW_SIZE_T and use the proper size for exporting the InnoDB variables. Remove some unnecessary indirection via export_vars, and bind some variables directly. dict_sys_t::rough_size(): Replaces dict_sys_get_size() and includes the hash table sizes. This is based on a contribution by Tony Liu from ServiceNow.
2019-07-03 16:31:20 +02:00
ulint buf_lru_flush_page_count;
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-25085: Simplify instrumentation for LRU eviction Let us add the status variable innodb_buffer_pool_pages_LRU_freed to monitor the number of pages that were freed by a buffer pool LRU eviction scan, without flushing. Also, let us simplify the monitor interface: MONITOR_LRU_BATCH_FLUSH_COUNT, MONITOR_LRU_BATCH_FLUSH_PAGES, MONITOR_LRU_BATCH_EVICT_COUNT, MONITOR_LRU_BATCH_EVICT_PAGES: Remove. MONITOR_LRU_BATCH_FLUSH_TOTAL_PAGE: Track buf_lru_flush_page_count (innodb_buffer_pool_pages_LRU_flushed). MONITOR_LRU_BATCH_EVICT_TOTAL_PAGE: Track buf_lru_freed_page_count (buffer_pool_pages_LRU_freed). Reviewed by: Vladislav Vaintroub
2021-03-09 07:29:38 +01:00
/** Number of pages freed without flushing. Protected by buf_pool.mutex. */
ulint buf_lru_freed_page_count;
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
/** Number of pages flushed. Protected by buf_pool.mutex. */
ulint buf_flush_page_count;
/** Flag indicating if the page_cleaner is in active state. */
TSAN: unprotected global variable Write of size 1 at 0x0000067abe08 by thread T3 (mutexes: write M1372): #0 buf_flush_page_cleaner() /storage/innobase/buf/buf0flu.cc:2366:29 (mariadbd+0x20cea7c) #1 void std::__invoke_impl<void, void (*)()>(std::__invoke_other, void (*&&)()) /usr/lib/gcc/x86_64-linux-gnu/11/../../../../include/c++/11/bits/invoke.h:61:14 (mariadbd+0x20c3a8a) #2 std::__invoke_result<void (*)()>::type std::__invoke<void (*)()>(void (*&&)()) /usr/lib/gcc/x86_64-linux-gnu/11/../../../../include/c++/11/bits/invoke.h:96:14 (mariadbd+0x20c399d) #3 void std::thread::_Invoker<std::tuple<void (*)()> >::_M_invoke<0ul>(std::_Index_tuple<0ul>) /usr/lib/gcc/x86_64-linux-gnu/11/../../../../include/c++/11/bits/std_thread.h:253:13 (mariadbd+0x20c3945) #4 std::thread::_Invoker<std::tuple<void (*)()> >::operator()() /usr/lib/gcc/x86_64-linux-gnu/11/../../../../include/c++/11/bits/std_thread.h:260:11 (mariadbd+0x20c38e5) #5 std::thread::_State_impl<std::thread::_Invoker<std::tuple<void (*)()> > >::_M_run() /usr/lib/gcc/x86_64-linux-gnu/11/../../../../include/c++/11/bits/std_thread.h:211:13 (mariadbd+0x20c37d9) #6 <null> <null> (libstdc++.so.6+0xd230f) Previous read of size 1 at 0x0000067abe08 by main thread: #0 logs_empty_and_mark_files_at_shutdown() /storage/innobase/log/log0log.cc:1094:6 (mariadbd+0x22eeff3) #1 innodb_shutdown() /storage/innobase/srv/srv0start.cc:1970:3 (mariadbd+0x256ffd6) #2 innobase_end(handlerton*, ha_panic_function) /storage/innobase/handler/ha_innodb.cc:4265:3 (mariadbd+0x1ee3fc4) #3 ha_finalize_handlerton(st_plugin_int*) /sql/handler.cc:595:5 (mariadbd+0xf7bac9) #4 plugin_deinitialize(st_plugin_int*, bool) /sql/sql_plugin.cc:1266:9 (mariadbd+0x1611789) #5 reap_plugins() /sql/sql_plugin.cc:1342:7 (mariadbd+0x160e17d) #6 plugin_shutdown() /sql/sql_plugin.cc:2050:7 (mariadbd+0x1611f42) #7 clean_up(bool) /sql/mysqld.cc:1923:3 (mariadbd+0xd67a4c) #8 unireg_abort /sql/mysqld.cc:1835:3 (mariadbd+0xd67605) #9 mysqld_main(int, char**) /sql/mysqld.cc:5741:7 (mariadbd+0xd6b36a) #10 main /sql/main.cc:34:10 (mariadbd+0xd661a8) Location is global 'buf_page_cleaner_is_active' of size 1 at 0x0000067abe08 (mariadbd+0x67abe08)
2021-09-08 22:51:52 +02:00
Atomic_relaxed<bool> buf_page_cleaner_is_active;
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
/** Factor for scan length to determine n_pages for intended oldest LSN
progress */
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
static constexpr ulint buf_flush_lsn_scan_factor = 3;
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
/** Average redo generation rate */
static lsn_t lsn_avg_rate = 0;
MDEV-25113: Introduce a page cleaner mode before 'furious flush' MDEV-23855 changed the way how the page cleaner is signaled by user threads. If a threshold is exceeded, a mini-transaction commit would invoke buf_flush_ahead() in order to initiate page flushing before all writers would eventually grind to halt in log_free_check(), waiting for the checkpoint age to reduce. However, buf_flush_ahead() would always initiate 'furious flushing', making the buf_flush_page_cleaner thread write innodb_io_capacity_max pages per batch, and sleeping no time between batches, until the limit LSN is reached. Because this could saturate the I/O subsystem, system throughput could significantly reduce during these 'furious flushing' spikes. With this change, we introduce a gentler version of flush-ahead, which would write innodb_io_capacity_max pages per second until the 'soft limit' is reached. buf_flush_ahead(): Add a parameter to specify whether furious flushing is requested. buf_flush_async_lsn: Similar to buf_flush_sync_lsn, a limit for the less intrusive flushing. buf_flush_page_cleaner(): Keep working until buf_flush_async_lsn has been reached. log_close(): Suppress a warning message in the event that a new log is being created during startup, when old logs did not exist. Return what type of page cleaning will be needed. mtr_t::finish_write(): Also when m_log.is_small(), invoke log_close(). Return what type of page cleaning will be needed. mtr_t::commit(): Invoke buf_flush_ahead() based on the return value of mtr_t::finish_write().
2021-06-23 12:13:16 +02:00
/** Target oldest_modification for the page cleaner background flushing;
writes are protected by buf_pool.flush_list_mutex */
static Atomic_relaxed<lsn_t> buf_flush_async_lsn;
/** Target oldest_modification for the page cleaner furious flushing;
writes are protected by buf_pool.flush_list_mutex */
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
static Atomic_relaxed<lsn_t> buf_flush_sync_lsn;
Merge InnoDB 5.7 from mysql-5.7.14. Contains also: MDEV-10549 mysqld: sql/handler.cc:2692: int handler::ha_index_first(uchar*): Assertion `table_share->tmp_table != NO_TMP_TABLE || m_lock_type != 2' failed. (branch bb-10.2-jan) Unlike MySQL, InnoDB still uses THR_LOCK in MariaDB MDEV-10548 Some of the debug sync waits do not work with InnoDB 5.7 (branch bb-10.2-jan) enable tests that were fixed in MDEV-10549 MDEV-10548 Some of the debug sync waits do not work with InnoDB 5.7 (branch bb-10.2-jan) fix main.innodb_mysql_sync - re-enable online alter for partitioned innodb tables
2016-09-06 08:43:16 +02:00
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
#ifdef UNIV_PFS_THREAD
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
mysql_pfs_key_t page_cleaner_thread_key;
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
#endif /* UNIV_PFS_THREAD */
MDEV-15058: Deprecate and ignore innodb_buffer_pool_instances Our benchmarking efforts indicate that the reasons for splitting the buf_pool in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 have mostly gone away, possibly as a result of mysql/mysql-server@ce6109ebfdedfdf185e391a0c97dc6d33867ed78 or similar work. Only in one write-heavy benchmark where the working set size is ten times the buffer pool size, the buf_pool->mutex would be less contended with 4 buffer pool instances than with 1 instance, in buf_page_io_complete(). That contention could be alleviated further by making more use of std::atomic and by splitting buf_pool_t::mutex further (MDEV-15053). We will deprecate and ignore the following parameters: innodb_buffer_pool_instances innodb_page_cleaners There will be only one buffer pool and one page cleaner task. In a number of INFORMATION_SCHEMA views, columns that indicated the buffer pool instance will be removed: information_schema.innodb_buffer_page.pool_id information_schema.innodb_buffer_page_lru.pool_id information_schema.innodb_buffer_pool_stats.pool_id information_schema.innodb_cmpmem.buffer_pool_instance information_schema.innodb_cmpmem_reset.buffer_pool_instance
2020-02-12 13:45:21 +01:00
/** Page cleaner structure */
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
static struct
{
/** total elapsed time in adaptive flushing, in seconds */
ulint flush_time;
/** number of adaptive flushing passes */
ulint flush_pass;
} page_cleaner;
MDEV-6936: Buffer pool list scan optimization Merged Facebook commit 617aef9f911d825e9053f3d611d0389e02031225 authored by Inaam Rana to InnoDB storage engine (not XtraDB) from https://github.com/facebook/mysql-5.6 WL#7047 - Optimize buffer pool list scans and related batch processing Reduce excessive scanning of pages when doing flush list batches. The fix is to introduce the concept of "Hazard Pointer", this reduces the time complexity of the scan from O(n*n) to O. The concept of hazard pointer is reversed in this work. Academically hazard pointer is a pointer that the thread working on it will declar such and as long as that thread is not done no other thread is allowe do anything with it. In this WL we declare the pointer as a hazard pointer and then if any thread attempts to work on it, it is allowed to do so but it has to a the hazard pointer to the next valid value. We use hazard pointer sol reverse traversal of lists within a buffer pool instance. Add an event to control the background flush thread. The background f thread wait has been converted to an os event timed wait so that it c signalled by threads that want to kick start a background flush when buffer pool is running low on free/dirty pages.
2014-11-06 12:17:11 +01:00
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
/* @} */
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
#ifdef UNIV_DEBUG
MDEV-15058: Deprecate and ignore innodb_buffer_pool_instances Our benchmarking efforts indicate that the reasons for splitting the buf_pool in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 have mostly gone away, possibly as a result of mysql/mysql-server@ce6109ebfdedfdf185e391a0c97dc6d33867ed78 or similar work. Only in one write-heavy benchmark where the working set size is ten times the buffer pool size, the buf_pool->mutex would be less contended with 4 buffer pool instances than with 1 instance, in buf_page_io_complete(). That contention could be alleviated further by making more use of std::atomic and by splitting buf_pool_t::mutex further (MDEV-15053). We will deprecate and ignore the following parameters: innodb_buffer_pool_instances innodb_page_cleaners There will be only one buffer pool and one page cleaner task. In a number of INFORMATION_SCHEMA views, columns that indicated the buffer pool instance will be removed: information_schema.innodb_buffer_page.pool_id information_schema.innodb_buffer_page_lru.pool_id information_schema.innodb_buffer_pool_stats.pool_id information_schema.innodb_cmpmem.buffer_pool_instance information_schema.innodb_cmpmem_reset.buffer_pool_instance
2020-02-12 13:45:21 +01:00
/** Validate the flush list. */
static void buf_flush_validate_low();
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-15058: Deprecate and ignore innodb_buffer_pool_instances Our benchmarking efforts indicate that the reasons for splitting the buf_pool in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 have mostly gone away, possibly as a result of mysql/mysql-server@ce6109ebfdedfdf185e391a0c97dc6d33867ed78 or similar work. Only in one write-heavy benchmark where the working set size is ten times the buffer pool size, the buf_pool->mutex would be less contended with 4 buffer pool instances than with 1 instance, in buf_page_io_complete(). That contention could be alleviated further by making more use of std::atomic and by splitting buf_pool_t::mutex further (MDEV-15053). We will deprecate and ignore the following parameters: innodb_buffer_pool_instances innodb_page_cleaners There will be only one buffer pool and one page cleaner task. In a number of INFORMATION_SCHEMA views, columns that indicated the buffer pool instance will be removed: information_schema.innodb_buffer_page.pool_id information_schema.innodb_buffer_page_lru.pool_id information_schema.innodb_buffer_pool_stats.pool_id information_schema.innodb_cmpmem.buffer_pool_instance information_schema.innodb_cmpmem_reset.buffer_pool_instance
2020-02-12 13:45:21 +01:00
/** Validates the flush list some of the time. */
static void buf_flush_validate_skip()
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
{
/** Try buf_flush_validate_low() every this many times */
# define BUF_FLUSH_VALIDATE_SKIP 23
/** The buf_flush_validate_low() call skip counter.
Use a signed type because of the race condition below. */
static int buf_flush_validate_count = BUF_FLUSH_VALIDATE_SKIP;
/* There is a race condition below, but it does not matter,
because this call is only for heuristic purposes. We want to
reduce the call frequency of the costly buf_flush_validate_low()
check in debug builds. */
if (--buf_flush_validate_count > 0) {
MDEV-15058: Deprecate and ignore innodb_buffer_pool_instances Our benchmarking efforts indicate that the reasons for splitting the buf_pool in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 have mostly gone away, possibly as a result of mysql/mysql-server@ce6109ebfdedfdf185e391a0c97dc6d33867ed78 or similar work. Only in one write-heavy benchmark where the working set size is ten times the buffer pool size, the buf_pool->mutex would be less contended with 4 buffer pool instances than with 1 instance, in buf_page_io_complete(). That contention could be alleviated further by making more use of std::atomic and by splitting buf_pool_t::mutex further (MDEV-15053). We will deprecate and ignore the following parameters: innodb_buffer_pool_instances innodb_page_cleaners There will be only one buffer pool and one page cleaner task. In a number of INFORMATION_SCHEMA views, columns that indicated the buffer pool instance will be removed: information_schema.innodb_buffer_page.pool_id information_schema.innodb_buffer_page_lru.pool_id information_schema.innodb_buffer_pool_stats.pool_id information_schema.innodb_cmpmem.buffer_pool_instance information_schema.innodb_cmpmem_reset.buffer_pool_instance
2020-02-12 13:45:21 +01:00
return;
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
}
buf_flush_validate_count = BUF_FLUSH_VALIDATE_SKIP;
MDEV-15058: Deprecate and ignore innodb_buffer_pool_instances Our benchmarking efforts indicate that the reasons for splitting the buf_pool in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 have mostly gone away, possibly as a result of mysql/mysql-server@ce6109ebfdedfdf185e391a0c97dc6d33867ed78 or similar work. Only in one write-heavy benchmark where the working set size is ten times the buffer pool size, the buf_pool->mutex would be less contended with 4 buffer pool instances than with 1 instance, in buf_page_io_complete(). That contention could be alleviated further by making more use of std::atomic and by splitting buf_pool_t::mutex further (MDEV-15053). We will deprecate and ignore the following parameters: innodb_buffer_pool_instances innodb_page_cleaners There will be only one buffer pool and one page cleaner task. In a number of INFORMATION_SCHEMA views, columns that indicated the buffer pool instance will be removed: information_schema.innodb_buffer_page.pool_id information_schema.innodb_buffer_page_lru.pool_id information_schema.innodb_buffer_pool_stats.pool_id information_schema.innodb_cmpmem.buffer_pool_instance information_schema.innodb_cmpmem_reset.buffer_pool_instance
2020-02-12 13:45:21 +01:00
buf_flush_validate_low();
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
}
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
#endif /* UNIV_DEBUG */
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-24278 InnoDB page cleaner keeps waking up on idle server The purpose of the InnoDB page cleaner subsystem is to write out modified pages from the buffer pool to data files. When the innodb_max_dirty_pages_pct_lwm is not exceeded or innodb_adaptive_flushing=ON decides not to write out anything, the page cleaner should keep sleeping indefinitely until the state of the system changes: a dirty page is added to the buffer pool such that the page cleaner would no longer be idle. buf_flush_page_cleaner(): Explicitly note when the page cleaner is idle. When that happens, use mysql_cond_wait() instead of mysql_cond_timedwait(). buf_flush_insert_into_flush_list(): Wake up the page cleaner if needed. innodb_max_dirty_pages_pct_update(), innodb_max_dirty_pages_pct_lwm_update(): Wake up the page cleaner just in case. Note: buf_flush_ahead(), buf_flush_wait_flushed() and shutdown are already waking up the page cleaner thread.
2020-11-25 15:09:47 +01:00
/** Wake up the page cleaner if needed */
MDEV-26819 SET GLOBAL innodb_max_dirty_pages_pct=0 occasionally fails to trigger writes innodb_max_dirty_pages_pct_update(), innodb_max_dirty_pages_pct_lwm_update(): Invoke buf_pool.page_cleaner_wakeup() in order to wake up buf_flush_page_cleaner. This allows the test innodb.page_cleaner to run without any occasional timeouts. The occasional hangs were introduced by commit 7b1252c03d7131754d9503560fe507b33ca1f8b4 (MDEV-24278).
2021-10-13 14:16:23 +02:00
void buf_pool_t::page_cleaner_wakeup()
MDEV-24278 InnoDB page cleaner keeps waking up on idle server The purpose of the InnoDB page cleaner subsystem is to write out modified pages from the buffer pool to data files. When the innodb_max_dirty_pages_pct_lwm is not exceeded or innodb_adaptive_flushing=ON decides not to write out anything, the page cleaner should keep sleeping indefinitely until the state of the system changes: a dirty page is added to the buffer pool such that the page cleaner would no longer be idle. buf_flush_page_cleaner(): Explicitly note when the page cleaner is idle. When that happens, use mysql_cond_wait() instead of mysql_cond_timedwait(). buf_flush_insert_into_flush_list(): Wake up the page cleaner if needed. innodb_max_dirty_pages_pct_update(), innodb_max_dirty_pages_pct_lwm_update(): Wake up the page cleaner just in case. Note: buf_flush_ahead(), buf_flush_wait_flushed() and shutdown are already waking up the page cleaner thread.
2020-11-25 15:09:47 +01:00
{
MDEV-24537 innodb_max_dirty_pages_pct_lwm=0 lost its special meaning In commit 3a9a3be1c64b14c05648e87ebe0f1dd96457de41 (MDEV-23855) some previous logic was replaced with the condition dirty_pct < srv_max_dirty_pages_pct_lwm, which caused the default value of the parameter innodb_max_dirty_pages_pct_lwm=0 to lose its special meaning: 'refer to innodb_max_dirty_pages_pct instead'. This implicit special meaning was visible in the function af_get_pct_for_dirty(), which was removed in commit f0c295e2de8c074c2ca72e19ff06e1d0e3ee6d2b (MDEV-24369). page_cleaner_flush_pages_recommendation(): Restore the special meaning that was removed in MDEV-24369. buf_flush_page_cleaner(): If srv_max_dirty_pages_pct_lwm==0.0, refer to srv_max_buf_pool_modified_pct. This fixes the observed performance regression due to excessive page flushing. buf_pool_t::page_cleaner_wakeup(): Revise the wakeup condition. innodb_init(): Do initialize srv_max_io_capacity in Mariabackup. It was previously constantly 0, which caused mariadb-backup --prepare to hang in buf_flush_sync(), making no progress.
2021-01-06 12:53:14 +01:00
if (!page_cleaner_idle())
return;
double dirty_pct= double(UT_LIST_GET_LEN(buf_pool.flush_list)) * 100.0 /
double(UT_LIST_GET_LEN(buf_pool.LRU) + UT_LIST_GET_LEN(buf_pool.free));
double pct_lwm= srv_max_dirty_pages_pct_lwm;
MDEV-24949: Enabling idle flushing (possible regression from MDEV-23855) - Currently page cleaner thread will stop flushing if dirty_pct < innodb_max_dirty_pages_pct_lwm. - If the server is not performing any activity then said resources/time could be used to flush the pending dirty pages and keep buffer pool clean for the next burst of the cycle. This flushing is called idle flushing. - flushing logic underwent a complete revamp in 10.5.7/8 and as part of the revamp idle flushing logic got removed. - New proposed logic of idle flushing is based on updated logic of the page cleaner that will enable idle flushing if - buf page cleaner is idle - there are dirty pages (< innodb_max_dirty_pages_pct_lwm) - server is not performing any activity Logic will kickstart the idle flushing bounded by innodb_io_capacity. (Thanks to Marko Makela for reviewing the patch and idea right from the its inception).
2021-02-23 09:21:30 +01:00
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
/* if pct_lwm != 0.0, adaptive flushing is enabled.
MDEV-24949: Enabling idle flushing (possible regression from MDEV-23855) - Currently page cleaner thread will stop flushing if dirty_pct < innodb_max_dirty_pages_pct_lwm. - If the server is not performing any activity then said resources/time could be used to flush the pending dirty pages and keep buffer pool clean for the next burst of the cycle. This flushing is called idle flushing. - flushing logic underwent a complete revamp in 10.5.7/8 and as part of the revamp idle flushing logic got removed. - New proposed logic of idle flushing is based on updated logic of the page cleaner that will enable idle flushing if - buf page cleaner is idle - there are dirty pages (< innodb_max_dirty_pages_pct_lwm) - server is not performing any activity Logic will kickstart the idle flushing bounded by innodb_io_capacity. (Thanks to Marko Makela for reviewing the patch and idea right from the its inception).
2021-02-23 09:21:30 +01:00
signal buf page cleaner thread
- if pct_lwm <= dirty_pct then it will invoke apdative flushing flow
- if pct_lwm > dirty_pct then it will invoke idle flushing flow.
idle_flushing:
dirty_pct < innodb_max_dirty_pages_pct_lwm so it could be an
idle flushing use-case.
Why is last_activity_count not updated always?
- let's first understand when is server activity count updated.
- it is updated on commit of a transaction trx_t::commit() and not
on adding a page to the flush list.
- page_cleaner_wakeup is called when a page is added to the flush list.
- now let's say the first user thread, updates the count from X -> Y but
is yet to commit the transaction (so activity count is still Y).
followup user threads will see the updated count as (Y) that is matching
the universal server activity count (Y), giving a false impression that
the server is idle.
How to avoid this?
- by allowing last_activity_count to updated when page-cleaner is made
active and has work to do. This ensures that the last_activity signal
is consumed by the page-cleaner before the next one is generated. */
MDEV-24537 innodb_max_dirty_pages_pct_lwm=0 lost its special meaning In commit 3a9a3be1c64b14c05648e87ebe0f1dd96457de41 (MDEV-23855) some previous logic was replaced with the condition dirty_pct < srv_max_dirty_pages_pct_lwm, which caused the default value of the parameter innodb_max_dirty_pages_pct_lwm=0 to lose its special meaning: 'refer to innodb_max_dirty_pages_pct instead'. This implicit special meaning was visible in the function af_get_pct_for_dirty(), which was removed in commit f0c295e2de8c074c2ca72e19ff06e1d0e3ee6d2b (MDEV-24369). page_cleaner_flush_pages_recommendation(): Restore the special meaning that was removed in MDEV-24369. buf_flush_page_cleaner(): If srv_max_dirty_pages_pct_lwm==0.0, refer to srv_max_buf_pool_modified_pct. This fixes the observed performance regression due to excessive page flushing. buf_pool_t::page_cleaner_wakeup(): Revise the wakeup condition. innodb_init(): Do initialize srv_max_io_capacity in Mariabackup. It was previously constantly 0, which caused mariadb-backup --prepare to hang in buf_flush_sync(), making no progress.
2021-01-06 12:53:14 +01:00
if ((pct_lwm != 0.0 && pct_lwm <= dirty_pct) ||
MDEV-24949: Enabling idle flushing (possible regression from MDEV-23855) - Currently page cleaner thread will stop flushing if dirty_pct < innodb_max_dirty_pages_pct_lwm. - If the server is not performing any activity then said resources/time could be used to flush the pending dirty pages and keep buffer pool clean for the next burst of the cycle. This flushing is called idle flushing. - flushing logic underwent a complete revamp in 10.5.7/8 and as part of the revamp idle flushing logic got removed. - New proposed logic of idle flushing is based on updated logic of the page cleaner that will enable idle flushing if - buf page cleaner is idle - there are dirty pages (< innodb_max_dirty_pages_pct_lwm) - server is not performing any activity Logic will kickstart the idle flushing bounded by innodb_io_capacity. (Thanks to Marko Makela for reviewing the patch and idea right from the its inception).
2021-02-23 09:21:30 +01:00
(pct_lwm != 0.0 && last_activity_count == srv_get_activity_count()) ||
MDEV-24537 innodb_max_dirty_pages_pct_lwm=0 lost its special meaning In commit 3a9a3be1c64b14c05648e87ebe0f1dd96457de41 (MDEV-23855) some previous logic was replaced with the condition dirty_pct < srv_max_dirty_pages_pct_lwm, which caused the default value of the parameter innodb_max_dirty_pages_pct_lwm=0 to lose its special meaning: 'refer to innodb_max_dirty_pages_pct instead'. This implicit special meaning was visible in the function af_get_pct_for_dirty(), which was removed in commit f0c295e2de8c074c2ca72e19ff06e1d0e3ee6d2b (MDEV-24369). page_cleaner_flush_pages_recommendation(): Restore the special meaning that was removed in MDEV-24369. buf_flush_page_cleaner(): If srv_max_dirty_pages_pct_lwm==0.0, refer to srv_max_buf_pool_modified_pct. This fixes the observed performance regression due to excessive page flushing. buf_pool_t::page_cleaner_wakeup(): Revise the wakeup condition. innodb_init(): Do initialize srv_max_io_capacity in Mariabackup. It was previously constantly 0, which caused mariadb-backup --prepare to hang in buf_flush_sync(), making no progress.
2021-01-06 12:53:14 +01:00
srv_max_buf_pool_modified_pct <= dirty_pct)
MDEV-24278 InnoDB page cleaner keeps waking up on idle server The purpose of the InnoDB page cleaner subsystem is to write out modified pages from the buffer pool to data files. When the innodb_max_dirty_pages_pct_lwm is not exceeded or innodb_adaptive_flushing=ON decides not to write out anything, the page cleaner should keep sleeping indefinitely until the state of the system changes: a dirty page is added to the buffer pool such that the page cleaner would no longer be idle. buf_flush_page_cleaner(): Explicitly note when the page cleaner is idle. When that happens, use mysql_cond_wait() instead of mysql_cond_timedwait(). buf_flush_insert_into_flush_list(): Wake up the page cleaner if needed. innodb_max_dirty_pages_pct_update(), innodb_max_dirty_pages_pct_lwm_update(): Wake up the page cleaner just in case. Note: buf_flush_ahead(), buf_flush_wait_flushed() and shutdown are already waking up the page cleaner thread.
2020-11-25 15:09:47 +01:00
{
page_cleaner_is_idle= false;
MDEV-23399 fixup: Use plain pthread_cond The condition variables that were introduced in commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) are never instrumented with PERFORMANCE_SCHEMA. Let us avoid the storage overhead and dead code.
2021-02-07 11:19:24 +01:00
pthread_cond_signal(&do_flush_list);
MDEV-24278 InnoDB page cleaner keeps waking up on idle server The purpose of the InnoDB page cleaner subsystem is to write out modified pages from the buffer pool to data files. When the innodb_max_dirty_pages_pct_lwm is not exceeded or innodb_adaptive_flushing=ON decides not to write out anything, the page cleaner should keep sleeping indefinitely until the state of the system changes: a dirty page is added to the buffer pool such that the page cleaner would no longer be idle. buf_flush_page_cleaner(): Explicitly note when the page cleaner is idle. When that happens, use mysql_cond_wait() instead of mysql_cond_timedwait(). buf_flush_insert_into_flush_list(): Wake up the page cleaner if needed. innodb_max_dirty_pages_pct_update(), innodb_max_dirty_pages_pct_lwm_update(): Wake up the page cleaner just in case. Note: buf_flush_ahead(), buf_flush_wait_flushed() and shutdown are already waking up the page cleaner thread.
2020-11-25 15:09:47 +01:00
}
}
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
inline void buf_pool_t::delete_from_flush_list_low(buf_page_t *bpage)
{
ut_ad(!fsp_is_system_temporary(bpage->id().space()));
mysql_mutex_assert_owner(&flush_list_mutex);
flush_hp.adjust(bpage);
UT_LIST_REMOVE(flush_list, bpage);
}
MDEV-15058: Deprecate and ignore innodb_buffer_pool_instances Our benchmarking efforts indicate that the reasons for splitting the buf_pool in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 have mostly gone away, possibly as a result of mysql/mysql-server@ce6109ebfdedfdf185e391a0c97dc6d33867ed78 or similar work. Only in one write-heavy benchmark where the working set size is ten times the buffer pool size, the buf_pool->mutex would be less contended with 4 buffer pool instances than with 1 instance, in buf_page_io_complete(). That contention could be alleviated further by making more use of std::atomic and by splitting buf_pool_t::mutex further (MDEV-15053). We will deprecate and ignore the following parameters: innodb_buffer_pool_instances innodb_page_cleaners There will be only one buffer pool and one page cleaner task. In a number of INFORMATION_SCHEMA views, columns that indicated the buffer pool instance will be removed: information_schema.innodb_buffer_page.pool_id information_schema.innodb_buffer_page_lru.pool_id information_schema.innodb_buffer_pool_stats.pool_id information_schema.innodb_cmpmem.buffer_pool_instance information_schema.innodb_cmpmem_reset.buffer_pool_instance
2020-02-12 13:45:21 +01:00
/** Insert a modified block into the flush list.
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
@param block modified block
@param lsn start LSN of the mini-transaction that modified the block */
void buf_pool_t::insert_into_flush_list(buf_block_t *block, lsn_t lsn)
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
{
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
mysql_mutex_assert_not_owner(&mutex);
mysql_mutex_assert_owner(&log_sys.flush_order_mutex);
ut_ad(lsn > 2);
ut_ad(!fsp_is_system_temporary(block->page.id().space()));
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
mysql_mutex_lock(&flush_list_mutex);
if (ut_d(const lsn_t old=) block->page.oldest_modification())
{
ut_ad(old == 1);
delete_from_flush_list_low(&block->page);
}
else
stat.flush_list_bytes+= block->physical_size();
ut_ad(stat.flush_list_bytes <= curr_pool_size);
block->page.set_oldest_modification(lsn);
MEM_CHECK_DEFINED(block->page.zip.data
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
? block->page.zip.data : block->page.frame,
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
block->physical_size());
UT_LIST_ADD_FIRST(flush_list, &block->page);
ut_d(buf_flush_validate_skip());
page_cleaner_wakeup();
mysql_mutex_unlock(&flush_list_mutex);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
}
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
/** Remove a block from flush_list.
@param bpage buffer pool page
@param clear whether to invoke buf_page_t::clear_oldest_modification() */
void buf_pool_t::delete_from_flush_list(buf_page_t *bpage, bool clear)
MDEV-12227 Defer writes to the InnoDB temporary tablespace The flushing of the InnoDB temporary tablespace is unnecessarily tied to the write-ahead redo logging and redo log checkpoints, which must be tied to the page writes of persistent tablespaces. Let us simply omit any pages of temporary tables from buf_pool.flush_list. In this way, log checkpoints will never incur any 'collateral damage' of writing out unmodified changes for temporary tables. After this change, pages of the temporary tablespace can only be written out by buf_flush_lists(n_pages,0) as part of LRU eviction. Hopefully, most of the time, that code will never be executed, and instead, the temporary pages will be evicted by buf_release_freed_page() without ever being written back to the temporary tablespace file. This should improve the efficiency of the checkpoint flushing and the buf_flush_page_cleaner thread. Reviewed by: Vladislav Vaintroub
2020-12-09 08:22:13 +01:00
{
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
delete_from_flush_list_low(bpage);
stat.flush_list_bytes-= bpage->physical_size();
MDEV-26010: Assertion lsn > 2 failed in buf_pool_t::get_oldest_modification In commit 22b62edaedddb1cabd5b855cdd39a5e90a5695a2 (MDEV-25113) we introduced a race condition. buf_LRU_free_page() would read buf_page_t::oldest_modification() as 0 and assume that buf_page_t::list can be used (for attaching the block to the buf_pool.free list). In the observed race condition, buf_pool_t::delete_from_flush_list() had cleared the field, and buf_pool_t::delete_from_flush_list_low() was executing concurrently with buf_LRU_block_free_non_file_page(), which resulted in buf_pool.flush_list.end becoming corrupted. buf_pool_t::delete_from_flush_list(), buf_flush_relocate_on_flush_list(): First remove the block from buf_pool.flush_list, and only then invoke buf_page_t::clear_oldest_modification(), to ensure that reading oldest_modification()==0 really implies that the block no longer is in buf_pool.flush_list.
2021-06-24 20:55:10 +02:00
if (clear)
bpage->clear_oldest_modification();
MDEV-12227 Defer writes to the InnoDB temporary tablespace The flushing of the InnoDB temporary tablespace is unnecessarily tied to the write-ahead redo logging and redo log checkpoints, which must be tied to the page writes of persistent tablespaces. Let us simply omit any pages of temporary tables from buf_pool.flush_list. In this way, log checkpoints will never incur any 'collateral damage' of writing out unmodified changes for temporary tables. After this change, pages of the temporary tablespace can only be written out by buf_flush_lists(n_pages,0) as part of LRU eviction. Hopefully, most of the time, that code will never be executed, and instead, the temporary pages will be evicted by buf_release_freed_page() without ever being written back to the temporary tablespace file. This should improve the efficiency of the checkpoint flushing and the buf_flush_page_cleaner thread. Reviewed by: Vladislav Vaintroub
2020-12-09 08:22:13 +01:00
#ifdef UNIV_DEBUG
buf_flush_validate_skip();
#endif /* UNIV_DEBUG */
}
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
/** Remove all dirty pages belonging to a given tablespace when we are
deleting the data file of that tablespace.
The pages still remain a part of LRU and are evicted from
the list as they age towards the tail of the LRU.
@param id tablespace identifier */
void buf_flush_remove_pages(ulint id)
{
const page_id_t first(id, 0), end(id + 1, 0);
ut_ad(id);
mysql_mutex_lock(&buf_pool.mutex);
for (;;)
{
bool deferred= false;
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
mysql_mutex_lock(&buf_pool.flush_list_mutex);
for (buf_page_t *bpage= UT_LIST_GET_LAST(buf_pool.flush_list); bpage; )
{
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
const auto s= bpage->state();
ut_ad(s >= buf_page_t::REMOVE_HASH);
ut_ad(s < buf_page_t::READ_FIX || s >= buf_page_t::WRITE_FIX);
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
buf_page_t *prev= UT_LIST_GET_PREV(list, bpage);
const page_id_t bpage_id(bpage->id());
if (bpage_id < first || bpage_id >= end);
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
else if (s >= buf_page_t::WRITE_FIX)
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
deferred= true;
else
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
buf_pool.delete_from_flush_list(bpage);
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
bpage= prev;
}
mysql_mutex_unlock(&buf_pool.flush_list_mutex);
if (!deferred)
break;
mysql_mutex_unlock(&buf_pool.mutex);
cleanup: os_thread_yield() -> std::this_thread::yield()
2021-03-16 21:34:55 +01:00
std::this_thread::yield();
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
mysql_mutex_lock(&buf_pool.mutex);
buf_flush_wait_batch_end(false);
}
mysql_mutex_unlock(&buf_pool.mutex);
}
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
/*******************************************************************//**
Relocates a buffer control block on the flush_list.
Note that it is assumed that the contents of bpage have already been
copied to dpage.
IMPORTANT: When this function is called bpage and dpage are not
exact copies of each other. For example, they both will have different
::state. Also the ::list pointers in dpage may be stale. We need to
use the current list node (bpage) to do the list manipulation because
the list pointers could have changed between the time that we copied
the contents of bpage to the dpage and the flush list manipulation
below. */
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
ATTRIBUTE_COLD
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
void
buf_flush_relocate_on_flush_list(
/*=============================*/
buf_page_t* bpage, /*!< in/out: control block being moved */
buf_page_t* dpage) /*!< in/out: destination block */
{
buf_page_t* prev;
MDEV-26200 buf_pool.flush_list corrupted by buffer pool resizing or ROW_FORMAT=COMPRESSED The lazy deletion of clean blocks from buf_pool.flush_list that was introduced in commit 6441bc614a99f5cd6357c8a23b9f583c56d0a90c (MDEV-25113) introduced a race condition around the function buf_flush_relocate_on_flush_list(). The test innodb_zip.wl5522_debug_zip as well as the buffer pool resizing tests would occasionally fail in debug builds due to buf_pool.flush_list.count disagreeing with the actual length of the doubly-linked list. The safe procedure for relocating a block in buf_pool.flush_list should be as follows, now that we implement lazy deletion from buf_pool.flush_list: 1. Acquire buf_pool.mutex. 2. Acquire the exclusive buf_pool.page_hash.latch. 3. Acquire buf_pool.flush_list_mutex. 4. Copy the block descriptor. 5. Invoke buf_flush_relocate_on_flush_list(). 6. Release buf_pool.flush_list_mutex. buf_flush_relocate_on_flush_list(): Assert that buf_pool.flush_list_mutex is being held. Invoke buf_page_t::oldest_modification() only once, using std::memory_order_relaxed, now that the mutex protects us. buf_LRU_free_page(), buf_LRU_block_remove_hashed(): Avoid an unlock-lock cycle on hash_lock. (We must not acquire hash_lock while already holding buf_pool.flush_list_mutex, because that could lead to a deadlock due to latching order violation.)
2021-07-22 07:34:49 +02:00
mysql_mutex_assert_owner(&buf_pool.flush_list_mutex);
MDEV-12227 Defer writes to the InnoDB temporary tablespace The flushing of the InnoDB temporary tablespace is unnecessarily tied to the write-ahead redo logging and redo log checkpoints, which must be tied to the page writes of persistent tablespaces. Let us simply omit any pages of temporary tables from buf_pool.flush_list. In this way, log checkpoints will never incur any 'collateral damage' of writing out unmodified changes for temporary tables. After this change, pages of the temporary tablespace can only be written out by buf_flush_lists(n_pages,0) as part of LRU eviction. Hopefully, most of the time, that code will never be executed, and instead, the temporary pages will be evicted by buf_release_freed_page() without ever being written back to the temporary tablespace file. This should improve the efficiency of the checkpoint flushing and the buf_flush_page_cleaner thread. Reviewed by: Vladislav Vaintroub
2020-12-09 08:22:13 +01:00
ut_ad(!fsp_is_system_temporary(bpage->id().space()));
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
MDEV-26200 buf_pool.flush_list corrupted by buffer pool resizing or ROW_FORMAT=COMPRESSED The lazy deletion of clean blocks from buf_pool.flush_list that was introduced in commit 6441bc614a99f5cd6357c8a23b9f583c56d0a90c (MDEV-25113) introduced a race condition around the function buf_flush_relocate_on_flush_list(). The test innodb_zip.wl5522_debug_zip as well as the buffer pool resizing tests would occasionally fail in debug builds due to buf_pool.flush_list.count disagreeing with the actual length of the doubly-linked list. The safe procedure for relocating a block in buf_pool.flush_list should be as follows, now that we implement lazy deletion from buf_pool.flush_list: 1. Acquire buf_pool.mutex. 2. Acquire the exclusive buf_pool.page_hash.latch. 3. Acquire buf_pool.flush_list_mutex. 4. Copy the block descriptor. 5. Invoke buf_flush_relocate_on_flush_list(). 6. Release buf_pool.flush_list_mutex. buf_flush_relocate_on_flush_list(): Assert that buf_pool.flush_list_mutex is being held. Invoke buf_page_t::oldest_modification() only once, using std::memory_order_relaxed, now that the mutex protects us. buf_LRU_free_page(), buf_LRU_block_remove_hashed(): Avoid an unlock-lock cycle on hash_lock. (We must not acquire hash_lock while already holding buf_pool.flush_list_mutex, because that could lead to a deadlock due to latching order violation.)
2021-07-22 07:34:49 +02:00
const lsn_t lsn = bpage->oldest_modification();
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
if (!lsn) {
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
return;
}
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
ut_ad(lsn == 1 || lsn > 2);
ut_ad(dpage->oldest_modification() == lsn);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-26200 buf_pool.flush_list corrupted by buffer pool resizing or ROW_FORMAT=COMPRESSED The lazy deletion of clean blocks from buf_pool.flush_list that was introduced in commit 6441bc614a99f5cd6357c8a23b9f583c56d0a90c (MDEV-25113) introduced a race condition around the function buf_flush_relocate_on_flush_list(). The test innodb_zip.wl5522_debug_zip as well as the buffer pool resizing tests would occasionally fail in debug builds due to buf_pool.flush_list.count disagreeing with the actual length of the doubly-linked list. The safe procedure for relocating a block in buf_pool.flush_list should be as follows, now that we implement lazy deletion from buf_pool.flush_list: 1. Acquire buf_pool.mutex. 2. Acquire the exclusive buf_pool.page_hash.latch. 3. Acquire buf_pool.flush_list_mutex. 4. Copy the block descriptor. 5. Invoke buf_flush_relocate_on_flush_list(). 6. Release buf_pool.flush_list_mutex. buf_flush_relocate_on_flush_list(): Assert that buf_pool.flush_list_mutex is being held. Invoke buf_page_t::oldest_modification() only once, using std::memory_order_relaxed, now that the mutex protects us. buf_LRU_free_page(), buf_LRU_block_remove_hashed(): Avoid an unlock-lock cycle on hash_lock. (We must not acquire hash_lock while already holding buf_pool.flush_list_mutex, because that could lead to a deadlock due to latching order violation.)
2021-07-22 07:34:49 +02:00
/* Important that we adjust the hazard pointer before removing
the bpage from the flush list. */
buf_pool.flush_hp.adjust(bpage);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-26200 buf_pool.flush_list corrupted by buffer pool resizing or ROW_FORMAT=COMPRESSED The lazy deletion of clean blocks from buf_pool.flush_list that was introduced in commit 6441bc614a99f5cd6357c8a23b9f583c56d0a90c (MDEV-25113) introduced a race condition around the function buf_flush_relocate_on_flush_list(). The test innodb_zip.wl5522_debug_zip as well as the buffer pool resizing tests would occasionally fail in debug builds due to buf_pool.flush_list.count disagreeing with the actual length of the doubly-linked list. The safe procedure for relocating a block in buf_pool.flush_list should be as follows, now that we implement lazy deletion from buf_pool.flush_list: 1. Acquire buf_pool.mutex. 2. Acquire the exclusive buf_pool.page_hash.latch. 3. Acquire buf_pool.flush_list_mutex. 4. Copy the block descriptor. 5. Invoke buf_flush_relocate_on_flush_list(). 6. Release buf_pool.flush_list_mutex. buf_flush_relocate_on_flush_list(): Assert that buf_pool.flush_list_mutex is being held. Invoke buf_page_t::oldest_modification() only once, using std::memory_order_relaxed, now that the mutex protects us. buf_LRU_free_page(), buf_LRU_block_remove_hashed(): Avoid an unlock-lock cycle on hash_lock. (We must not acquire hash_lock while already holding buf_pool.flush_list_mutex, because that could lead to a deadlock due to latching order violation.)
2021-07-22 07:34:49 +02:00
prev = UT_LIST_GET_PREV(list, bpage);
UT_LIST_REMOVE(buf_pool.flush_list, bpage);
MDEV-26010: Assertion lsn > 2 failed in buf_pool_t::get_oldest_modification In commit 22b62edaedddb1cabd5b855cdd39a5e90a5695a2 (MDEV-25113) we introduced a race condition. buf_LRU_free_page() would read buf_page_t::oldest_modification() as 0 and assume that buf_page_t::list can be used (for attaching the block to the buf_pool.free list). In the observed race condition, buf_pool_t::delete_from_flush_list() had cleared the field, and buf_pool_t::delete_from_flush_list_low() was executing concurrently with buf_LRU_block_free_non_file_page(), which resulted in buf_pool.flush_list.end becoming corrupted. buf_pool_t::delete_from_flush_list(), buf_flush_relocate_on_flush_list(): First remove the block from buf_pool.flush_list, and only then invoke buf_page_t::clear_oldest_modification(), to ensure that reading oldest_modification()==0 really implies that the block no longer is in buf_pool.flush_list.
2021-06-24 20:55:10 +02:00
MDEV-26200 buf_pool.flush_list corrupted by buffer pool resizing or ROW_FORMAT=COMPRESSED The lazy deletion of clean blocks from buf_pool.flush_list that was introduced in commit 6441bc614a99f5cd6357c8a23b9f583c56d0a90c (MDEV-25113) introduced a race condition around the function buf_flush_relocate_on_flush_list(). The test innodb_zip.wl5522_debug_zip as well as the buffer pool resizing tests would occasionally fail in debug builds due to buf_pool.flush_list.count disagreeing with the actual length of the doubly-linked list. The safe procedure for relocating a block in buf_pool.flush_list should be as follows, now that we implement lazy deletion from buf_pool.flush_list: 1. Acquire buf_pool.mutex. 2. Acquire the exclusive buf_pool.page_hash.latch. 3. Acquire buf_pool.flush_list_mutex. 4. Copy the block descriptor. 5. Invoke buf_flush_relocate_on_flush_list(). 6. Release buf_pool.flush_list_mutex. buf_flush_relocate_on_flush_list(): Assert that buf_pool.flush_list_mutex is being held. Invoke buf_page_t::oldest_modification() only once, using std::memory_order_relaxed, now that the mutex protects us. buf_LRU_free_page(), buf_LRU_block_remove_hashed(): Avoid an unlock-lock cycle on hash_lock. (We must not acquire hash_lock while already holding buf_pool.flush_list_mutex, because that could lead to a deadlock due to latching order violation.)
2021-07-22 07:34:49 +02:00
bpage->clear_oldest_modification();
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
if (lsn == 1) {
MDEV-26017 fixup buf_flush_relocate_on_flush_list(): Use dpage->physical_size() because bpage->zip.ssize may already have been zeroed in page_zip_set_size() invoked by buf_pool_t::realloc(). This would cause occasional failures of the test innodb.innodb_buffer_pool_resize, which creates a ROW_FORMAT=COMPRESSED table.
2021-07-03 13:52:04 +02:00
buf_pool.stat.flush_list_bytes -= dpage->physical_size();
MDEV-26010 fixup: Use acquire/release memory order In commit 5f22511e35674ecc376f4c56217ee2d78f92c772 we depend on Total Store Ordering. For correct operation on ISAs that implement weaker memory ordering, we must explicitly use release/acquire stores and loads on buf_page_t::oldest_modification_ to prevent a race condition when buf_page_t::list does not happen to be on the same cache line. buf_page_t::clear_oldest_modification(): Assert that the block is not in buf_pool.flush_list, and use std::memory_order_release. buf_page_t::oldest_modification_acquire(): Read oldest_modification_ with std::memory_order_acquire. In this way, if the return value is 0, the caller may safely assume that it will not observe the buf_page_t as being in buf_pool.flush_list, even if it is not holding buf_pool.flush_list_mutex. buf_flush_relocate_on_flush_list(), buf_LRU_free_page(): Invoke buf_page_t::oldest_modification_acquire().
2021-06-26 10:16:40 +02:00
dpage->list.prev = nullptr;
dpage->list.next = nullptr;
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
dpage->clear_oldest_modification();
} else if (prev) {
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
ut_ad(prev->oldest_modification());
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
UT_LIST_INSERT_AFTER(buf_pool.flush_list, prev, dpage);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
} else {
MDEV-21962 Allocate buf_pool statically Thanks to MDEV-15058, there is only one InnoDB buffer pool. Allocating buf_pool statically removes one level of pointer indirection and makes code more readable, and removes the awkward initialization of some buf_pool members. While doing this, we will also declare some buf_pool_t data members private and replace some functions with member functions. This is mostly affecting buffer pool resizing. This is not aiming to be a complete rewrite of buf_pool_t to a proper class. Most of the buffer pool interface, such as buf_page_get_gen(), will remain in the C programming style for now. buf_pool_t::withdrawing: Replaces buf_pool_withdrawing. buf_pool_t::withdraw_clock_: Replaces buf_withdraw_clock. buf_pool_t::create(): Repalces buf_pool_init(). buf_pool_t::close(): Replaces buf_pool_free(). buf_bool_t::will_be_withdrawn(): Replaces buf_block_will_be_withdrawn(), buf_frame_will_be_withdrawn(). buf_pool_t::clear_hash_index(): Replaces buf_pool_clear_hash_index(). buf_pool_t::get_n_pages(): Replaces buf_pool_get_n_pages(). buf_pool_t::validate(): Replaces buf_validate(). buf_pool_t::print(): Replaces buf_print(). buf_pool_t::block_from_ahi(): Replaces buf_block_from_ahi(). buf_pool_t::is_block_field(): Replaces buf_pointer_is_block_field(). buf_pool_t::is_block_mutex(): Replaces buf_pool_is_block_mutex(). buf_pool_t::is_block_lock(): Replaces buf_pool_is_block_lock(). buf_pool_t::is_obsolete(): Replaces buf_pool_is_obsolete(). buf_pool_t::io_buf: Make default-constructible. buf_pool_t::io_buf::create(): Delayed 'constructor' buf_pool_t::io_buf::close(): Early 'destructor' HazardPointer: Make default-constructible. Define all member functions inline, also for derived classes.
2020-03-18 20:48:00 +01:00
UT_LIST_ADD_FIRST(buf_pool.flush_list, dpage);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
}
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
ut_d(buf_flush_validate_low());
}
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
/** Note that a block is no longer dirty, while not removing
it from buf_pool.flush_list */
inline void buf_page_t::write_complete(bool temporary)
{
ut_ad(temporary == fsp_is_system_temporary(id().space()));
if (temporary)
{
ut_ad(oldest_modification() == 2);
oldest_modification_= 0;
}
else
{
/* We use release memory order to guarantee that callers of
oldest_modification_acquire() will observe the block as
being detached from buf_pool.flush_list, after reading the value 0. */
ut_ad(oldest_modification() > 2);
oldest_modification_.store(1, std::memory_order_release);
}
const auto s= state();
ut_ad(s >= WRITE_FIX);
zip.fix.fetch_sub((s >= WRITE_FIX_REINIT)
? (WRITE_FIX_REINIT - UNFIXED)
: (WRITE_FIX - UNFIXED));
lock.u_unlock(true);
}
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
/** Complete write of a file page from buf_pool.
MDEV-16264 fixup: Clean up asynchronous I/O os_aio_userdata_t: Remove. It was basically duplicating IORequest. buf_page_write_complete(): Take only IORequest as a parameter. os_aio_func(), pfs_os_aio_func(): Replaced with os_aio() that has no redundant parameters. There is only one caller, so there is no point to pass __FILE__, __LINE__ as a parameter.
2020-10-23 09:49:03 +02:00
@param request write request */
void buf_page_write_complete(const IORequest &request)
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
{
ut_ad(request.is_write());
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
ut_ad(!srv_read_only_mode);
MDEV-16264 fixup: Clean up asynchronous I/O os_aio_userdata_t: Remove. It was basically duplicating IORequest. buf_page_write_complete(): Take only IORequest as a parameter. os_aio_func(), pfs_os_aio_func(): Replaced with os_aio() that has no redundant parameters. There is only one caller, so there is no point to pass __FILE__, __LINE__ as a parameter.
2020-10-23 09:49:03 +02:00
buf_page_t *bpage= request.bpage;
ut_ad(bpage);
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
const auto state= bpage->state();
/* io-fix can only be cleared by buf_page_t::write_complete()
and buf_page_t::read_complete() */
ut_ad(state >= buf_page_t::WRITE_FIX);
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
ut_ad(!buf_dblwr.is_inside(bpage->id()));
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
ut_ad(request.node->space->id == bpage->id().space());
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
if (state < buf_page_t::WRITE_FIX_REINIT &&
request.node->space->use_doublewrite())
MDEV-16264 fixup: Clean up asynchronous I/O os_aio_userdata_t: Remove. It was basically duplicating IORequest. buf_page_write_complete(): Take only IORequest as a parameter. os_aio_func(), pfs_os_aio_func(): Replaced with os_aio() that has no redundant parameters. There is only one caller, so there is no point to pass __FILE__, __LINE__ as a parameter.
2020-10-23 09:49:03 +02:00
{
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
ut_ad(request.node->space != fil_system.temp_space);
buf_dblwr.write_completed();
MDEV-16264 fixup: Clean up asynchronous I/O os_aio_userdata_t: Remove. It was basically duplicating IORequest. buf_page_write_complete(): Take only IORequest as a parameter. os_aio_func(), pfs_os_aio_func(): Replaced with os_aio() that has no redundant parameters. There is only one caller, so there is no point to pass __FILE__, __LINE__ as a parameter.
2020-10-23 09:49:03 +02:00
}
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
MDEV-27058: Move buf_page_t::slot to IORequest::slot MDEV-23855 and MDEV-23399 already moved some transient data fields from buffer pool page descriptors to IORequest, but the write buffer of PAGE_COMPRESSED or ENCRYPTED tables was missed. Since is only needed during asynchronous page write requests, it belongs to IORequest.
2021-11-16 17:16:53 +01:00
if (request.slot)
request.slot->release();
MDEV-11802 innodb.innodb_bug14676111 fails The function trx_purge_stop() was calling os_event_reset(purge_sys->event) before calling rw_lock_x_lock(&purge_sys->latch). The os_event_set() call in srv_purge_coordinator_suspend() is protected by that X-latch. It would seem a good idea to consistently protect both os_event_set() and os_event_reset() calls with a common mutex or rw-lock in those cases where os_event_set() and os_event_reset() are used like condition variables, tied to changes of shared state. For each os_event_t, we try to document the mutex or rw-lock that is being used. For some events, frequent calls to os_event_set() seem to try to avoid hangs. Some events are never waited for infinitely, only timed waits, and os_event_set() is used for early termination of these waits. os_aio_simulated_put_read_threads_to_sleep(): Define as a null macro on other systems than Windows. TODO: remove this altogether and disable innodb_use_native_aio on Windows. os_aio_segment_wait_events[]: Initialize only if innodb_use_native_aio=0. log_write_flush_to_disk_low(): Invoke log_mutex_enter() at the end, to avoid race conditions when changing the system state. (No potential race condition existed before MySQL 5.7.)
2017-02-17 09:32:21 +01:00
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
if (UNIV_UNLIKELY(MONITOR_IS_ON(MONITOR_MODULE_BUF_PAGE)))
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
buf_page_monitor(*bpage, false);
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
DBUG_PRINT("ib_buf", ("write page %u:%u",
bpage->id().space(), bpage->id().page_no()));
MDEV-12227 Defer writes to the InnoDB temporary tablespace The flushing of the InnoDB temporary tablespace is unnecessarily tied to the write-ahead redo logging and redo log checkpoints, which must be tied to the page writes of persistent tablespaces. Let us simply omit any pages of temporary tables from buf_pool.flush_list. In this way, log checkpoints will never incur any 'collateral damage' of writing out unmodified changes for temporary tables. After this change, pages of the temporary tablespace can only be written out by buf_flush_lists(n_pages,0) as part of LRU eviction. Hopefully, most of the time, that code will never be executed, and instead, the temporary pages will be evicted by buf_release_freed_page() without ever being written back to the temporary tablespace file. This should improve the efficiency of the checkpoint flushing and the buf_flush_page_cleaner thread. Reviewed by: Vladislav Vaintroub
2020-12-09 08:22:13 +01:00
const bool temp= fsp_is_system_temporary(bpage->id().space());
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
mysql_mutex_lock(&buf_pool.mutex);
MDEV-26445 innodb_undo_log_truncate is unnecessarily slow trx_purge_truncate_history(): Do not force a write of the undo tablespace that is being truncated. Instead, prevent page writes by acquiring an exclusive latch on all dirty pages of the tablespace. fseg_create(): Relax an assertion that could fail if a dirty undo page is being initialized during undo tablespace truncation (and trx_purge_truncate_history() already acquired an exclusive latch on it). fsp_page_create(): If we are truncating a tablespace, try to reuse a page that we may have already latched exclusively (because it was in buf_pool.flush_list). To some extent, this helps the test innodb.undo_truncate,16k to avoid running out of buffer pool. mtr_t::commit_shrink(): Mark as clean all pages that are outside the new bounds of the tablespace, and only add the newly reinitialized pages to the buf_pool.flush_list. buf_page_create(): Do not unnecessarily invoke change buffer merge on undo tablespaces. buf_page_t::clear_oldest_modification(bool temporary): Move some assertions to the caller buf_page_write_complete(). innodb.undo_truncate: Use a bigger innodb_buffer_pool_size=24M. On my system, it would otherwise hang 1 out of 1547 attempts (on the 40th repeat of innodb.undo_truncate,16k). Other page sizes were not affected.
2021-09-24 07:24:03 +02:00
mysql_mutex_assert_not_owner(&buf_pool.flush_list_mutex);
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
buf_pool.stat.n_pages_written++;
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
bpage->write_complete(temp);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
if (request.is_LRU())
MDEV-26004 Excessive wait times in buf_LRU_get_free_block() buf_LRU_get_free_block(): Initially wait for a single block to be freed, signaled by buf_pool.done_free. Only if that fails and no LRU eviction flushing batch is already running, we initiate a flushing batch that should serve all threads that are currently waiting in buf_LRU_get_free_block(). Note: In an extreme case, this may introduce a performance regression at larger numbers of connections. We observed this in sysbench oltp_update_index with 512MiB buffer pool, 4GiB of data on fast NVMe, and 1000 concurrent connections, on a 20-thread CPU. The contention point appears to be buf_pool.mutex, and the improvement would turn into a regression somewhere beyond 32 concurrent connections. On slower storage, such regression was not observed; instead, the throughput was improving and maximum latency was reduced. The excessive waits were pointed out by Vladislav Vaintroub.
2021-06-24 10:01:18 +02:00
{
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
buf_LRU_free_page(bpage, true);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
ut_ad(buf_pool.n_flush_LRU_);
if (!--buf_pool.n_flush_LRU_)
MDEV-26004 Excessive wait times in buf_LRU_get_free_block() buf_LRU_get_free_block(): Initially wait for a single block to be freed, signaled by buf_pool.done_free. Only if that fails and no LRU eviction flushing batch is already running, we initiate a flushing batch that should serve all threads that are currently waiting in buf_LRU_get_free_block(). Note: In an extreme case, this may introduce a performance regression at larger numbers of connections. We observed this in sysbench oltp_update_index with 512MiB buffer pool, 4GiB of data on fast NVMe, and 1000 concurrent connections, on a 20-thread CPU. The contention point appears to be buf_pool.mutex, and the improvement would turn into a regression somewhere beyond 32 concurrent connections. On slower storage, such regression was not observed; instead, the throughput was improving and maximum latency was reduced. The excessive waits were pointed out by Vladislav Vaintroub.
2021-06-24 10:01:18 +02:00
{
MDEV-23399 fixup: Use plain pthread_cond The condition variables that were introduced in commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) are never instrumented with PERFORMANCE_SCHEMA. Let us avoid the storage overhead and dead code.
2021-02-07 11:19:24 +01:00
pthread_cond_broadcast(&buf_pool.done_flush_LRU);
MDEV-26004 Excessive wait times in buf_LRU_get_free_block() buf_LRU_get_free_block(): Initially wait for a single block to be freed, signaled by buf_pool.done_free. Only if that fails and no LRU eviction flushing batch is already running, we initiate a flushing batch that should serve all threads that are currently waiting in buf_LRU_get_free_block(). Note: In an extreme case, this may introduce a performance regression at larger numbers of connections. We observed this in sysbench oltp_update_index with 512MiB buffer pool, 4GiB of data on fast NVMe, and 1000 concurrent connections, on a 20-thread CPU. The contention point appears to be buf_pool.mutex, and the improvement would turn into a regression somewhere beyond 32 concurrent connections. On slower storage, such regression was not observed; instead, the throughput was improving and maximum latency was reduced. The excessive waits were pointed out by Vladislav Vaintroub.
2021-06-24 10:01:18 +02:00
pthread_cond_signal(&buf_pool.done_free);
}
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
}
else
{
MDEV-26004 Excessive wait times in buf_LRU_get_free_block() buf_LRU_get_free_block(): Initially wait for a single block to be freed, signaled by buf_pool.done_free. Only if that fails and no LRU eviction flushing batch is already running, we initiate a flushing batch that should serve all threads that are currently waiting in buf_LRU_get_free_block(). Note: In an extreme case, this may introduce a performance regression at larger numbers of connections. We observed this in sysbench oltp_update_index with 512MiB buffer pool, 4GiB of data on fast NVMe, and 1000 concurrent connections, on a 20-thread CPU. The contention point appears to be buf_pool.mutex, and the improvement would turn into a regression somewhere beyond 32 concurrent connections. On slower storage, such regression was not observed; instead, the throughput was improving and maximum latency was reduced. The excessive waits were pointed out by Vladislav Vaintroub.
2021-06-24 10:01:18 +02:00
ut_ad(!temp);
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
ut_ad(buf_pool.n_flush_list_);
if (!--buf_pool.n_flush_list_)
MDEV-23399 fixup: Use plain pthread_cond The condition variables that were introduced in commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) are never instrumented with PERFORMANCE_SCHEMA. Let us avoid the storage overhead and dead code.
2021-02-07 11:19:24 +01:00
pthread_cond_broadcast(&buf_pool.done_flush_list);
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
}
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
mysql_mutex_unlock(&buf_pool.mutex);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
}
MDEV-21132 Remove buf_page_t::newest_modification At each mini-transaction commit, the log sequence number of the mini-transaction must be written to each modified page, so that it will be available in the FIL_PAGE_LSN field when the page is being read in crash recovery. InnoDB was unnecessarily allocating redundant storage for the field, in buf_page_t::newest_modification. Let us access FIL_PAGE_LSN directly. Furthermore, on ALTER TABLE...IMPORT TABLESPACE, let us write 0 to FIL_PAGE_LSN instead of using log_sys.lsn. buf_flush_init_for_writing(), buf_flush_update_zip_checksum(), fil_encrypt_buf_for_full_crc32(), fil_encrypt_buf(), fil_space_encrypt(): Remove the parameter lsn. buf_page_get_newest_modification(): Merge with the only caller. buf_tmp_reserve_compression_buf(), buf_tmp_page_encrypt(), buf_page_encrypt(): Define static in the same compilation unit with the only caller. PageConverter::m_current_lsn: Remove. Write 0 to FIL_PAGE_LSN on ALTER TABLE...IMPORT TABLESPACE.
2019-11-25 08:39:51 +01:00
/** Calculate a ROW_FORMAT=COMPRESSED page checksum and update the page.
MDEV-18644: Support full_crc32 for page_compressed This is a follow-up task to MDEV-12026, which introduced innodb_checksum_algorithm=full_crc32 and a simpler page format. MDEV-12026 did not enable full_crc32 for page_compressed tables, which we will be doing now. This is joint work with Thirunarayanan Balathandayuthapani. For innodb_checksum_algorithm=full_crc32 we change the page_compressed format as follows: FIL_PAGE_TYPE: The most significant bit will be set to indicate page_compressed format. The least significant bits will contain the compressed page size, rounded up to a multiple of 256 bytes. The checksum will be stored in the last 4 bytes of the page (whether it is the full page or a page_compressed page whose size is determined by FIL_PAGE_TYPE), covering all preceding bytes of the page. If encryption is used, then the page will be encrypted between compression and computing the checksum. For page_compressed, FIL_PAGE_LSN will not be repeated at the end of the page. FSP_SPACE_FLAGS (already implemented as part of MDEV-12026): We will store the innodb_compression_algorithm that may be used to compress pages. Previously, the choice of algorithm was written to each compressed data page separately, and one would be unable to know in advance which compression algorithm(s) are used. fil_space_t::full_crc32_page_compressed_len(): Determine if the page_compressed algorithm of the tablespace needs to know the exact length of the compressed data. If yes, we will reserve and write an extra byte for this right before the checksum. buf_page_is_compressed(): Determine if a page uses page_compressed (in any innodb_checksum_algorithm). fil_page_decompress(): Pass also fil_space_t::flags so that the format can be determined. buf_page_is_zeroes(): Check if a page is full of zero bytes. buf_page_full_crc32_is_corrupted(): Renamed from buf_encrypted_full_crc32_page_is_corrupted(). For full_crc32, we always simply validate the checksum to the page contents, while the physical page size is explicitly specified by an unencrypted part of the page header. buf_page_full_crc32_size(): Determine the size of a full_crc32 page. buf_dblwr_check_page_lsn(): Make this a debug-only function, because it involves potentially costly lookups of fil_space_t. create_table_info_t::check_table_options(), ha_innobase::check_if_supported_inplace_alter(): Do allow the creation of SPATIAL INDEX with full_crc32 also when page_compressed is used. commit_cache_norebuild(): Preserve the compression algorithm when updating the page_compression_level. dict_tf_to_fsp_flags(): Set the flags for page compression algorithm. FIXME: Maybe there should be a table option page_compression_algorithm and a session variable to back it?
2019-03-18 13:08:43 +01:00
@param[in,out] page page to update
MDEV-21132 Remove buf_page_t::newest_modification At each mini-transaction commit, the log sequence number of the mini-transaction must be written to each modified page, so that it will be available in the FIL_PAGE_LSN field when the page is being read in crash recovery. InnoDB was unnecessarily allocating redundant storage for the field, in buf_page_t::newest_modification. Let us access FIL_PAGE_LSN directly. Furthermore, on ALTER TABLE...IMPORT TABLESPACE, let us write 0 to FIL_PAGE_LSN instead of using log_sys.lsn. buf_flush_init_for_writing(), buf_flush_update_zip_checksum(), fil_encrypt_buf_for_full_crc32(), fil_encrypt_buf(), fil_space_encrypt(): Remove the parameter lsn. buf_page_get_newest_modification(): Merge with the only caller. buf_tmp_reserve_compression_buf(), buf_tmp_page_encrypt(), buf_page_encrypt(): Define static in the same compilation unit with the only caller. PageConverter::m_current_lsn: Remove. Write 0 to FIL_PAGE_LSN on ALTER TABLE...IMPORT TABLESPACE.
2019-11-25 08:39:51 +01:00
@param[in] size compressed page size */
void buf_flush_update_zip_checksum(buf_frame_t *page, ulint size)
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
{
MDEV-21132 Remove buf_page_t::newest_modification At each mini-transaction commit, the log sequence number of the mini-transaction must be written to each modified page, so that it will be available in the FIL_PAGE_LSN field when the page is being read in crash recovery. InnoDB was unnecessarily allocating redundant storage for the field, in buf_page_t::newest_modification. Let us access FIL_PAGE_LSN directly. Furthermore, on ALTER TABLE...IMPORT TABLESPACE, let us write 0 to FIL_PAGE_LSN instead of using log_sys.lsn. buf_flush_init_for_writing(), buf_flush_update_zip_checksum(), fil_encrypt_buf_for_full_crc32(), fil_encrypt_buf(), fil_space_encrypt(): Remove the parameter lsn. buf_page_get_newest_modification(): Merge with the only caller. buf_tmp_reserve_compression_buf(), buf_tmp_page_encrypt(), buf_page_encrypt(): Define static in the same compilation unit with the only caller. PageConverter::m_current_lsn: Remove. Write 0 to FIL_PAGE_LSN on ALTER TABLE...IMPORT TABLESPACE.
2019-11-25 08:39:51 +01:00
ut_ad(size > 0);
mach_write_to_4(page + FIL_PAGE_SPACE_OR_CHKSUM,
MDEV-25105 Remove innodb_checksum_algorithm values none,innodb,... Historically, InnoDB supported a buggy page checksum algorithm that did not compute a checksum over the full page. Later, well before MySQL 4.1 introduced .ibd files and the innodb_file_per_table option, the algorithm was corrected and the first 4 bytes of each page were redefined to be a checksum. The original checksum was so slow that an option to disable page checksum was introduced for benchmarketing purposes. The Intel Nehalem microarchitecture introduced the SSE4.2 instruction set extension, which includes instructions for faster computation of CRC-32C. In MySQL 5.6 (and MariaDB 10.0), innodb_checksum_algorithm=crc32 was implemented to make of that. As that option was changed to be the default in MySQL 5.7, a bug was found on big-endian platforms and some work-around code was added to weaken that checksum further. MariaDB disables that work-around by default since MDEV-17958. Later, SIMD-accelerated CRC-32C has been implemented in MariaDB for POWER and ARM and also for IA-32/AMD64, making use of carry-less multiplication where available. Long story short, innodb_checksum_algorithm=crc32 is faster and more secure than the pre-MySQL 5.6 checksum, called innodb_checksum_algorithm=innodb. It should have removed any need to use innodb_checksum_algorithm=none. The setting innodb_checksum_algorithm=crc32 is the default in MySQL 5.7 and MariaDB Server 10.2, 10.3, 10.4. In MariaDB 10.5, MDEV-19534 made innodb_checksum_algorithm=full_crc32 the default. It is even faster and more secure. The default settings in MariaDB do allow old data files to be read, no matter if a worse checksum algorithm had been used. (Unfortunately, before innodb_checksum_algorithm=full_crc32, the data files did not identify which checksum algorithm is being used.) The non-default settings innodb_checksum_algorithm=strict_crc32 or innodb_checksum_algorithm=strict_full_crc32 would only allow CRC-32C checksums. The incompatibility with old data files is why they are not the default. The newest server not to support innodb_checksum_algorithm=crc32 were MySQL 5.5 and MariaDB 5.5. Both have reached their end of life. A valid reason for using innodb_checksum_algorithm=innodb could have been the ability to downgrade. If it is really needed, data files can be converted with an older version of the innochecksum utility. Because there is no good reason to allow data files to be written with insecure checksums, we will reject those option values: innodb_checksum_algorithm=none innodb_checksum_algorithm=innodb innodb_checksum_algorithm=strict_none innodb_checksum_algorithm=strict_innodb Furthermore, the following innochecksum options will be removed, because only strict crc32 will be supported: innochecksum --strict-check=crc32 innochecksum -C crc32 innochecksum --write=crc32 innochecksum -w crc32 If a user wishes to convert a data file to use a different checksum (so that it might be used with the no-longer-supported MySQL 5.5 or MariaDB 5.5, which do not support IMPORT TABLESPACE nor system tablespace format changes that were made in MariaDB 10.3), then the innochecksum tool from MariaDB 10.2, 10.3, 10.4, 10.5 or MySQL 5.7 can be used. Reviewed by: Thirunarayanan Balathandayuthapani
2021-03-11 10:56:35 +01:00
page_zip_calc_checksum(page, size, false));
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
}
MDEV-12026: Implement innodb_checksum_algorithm=full_crc32 MariaDB data-at-rest encryption (innodb_encrypt_tables) had repurposed the same unused data field that was repurposed in MySQL 5.7 (and MariaDB 10.2) for the Split Sequence Number (SSN) field of SPATIAL INDEX. Because of this, MariaDB was unable to support encryption on SPATIAL INDEX pages. Furthermore, InnoDB page checksums skipped some bytes, and there are multiple variations and checksum algorithms. By default, InnoDB accepts all variations of all algorithms that ever existed. This unnecessarily weakens the page checksums. We hereby introduce two more innodb_checksum_algorithm variants (full_crc32, strict_full_crc32) that are special in a way: When either setting is active, newly created data files will carry a flag (fil_space_t::full_crc32()) that indicates that all pages of the file will use a full CRC-32C checksum over the entire page contents (excluding the bytes where the checksum is stored, at the very end of the page). Such files will always use that checksum, no matter what the parameter innodb_checksum_algorithm is assigned to. For old files, the old checksum algorithms will continue to be used. The value strict_full_crc32 will be equivalent to strict_crc32 and the value full_crc32 will be equivalent to crc32. ROW_FORMAT=COMPRESSED tables will only use the old format. These tables do not support new features, such as larger innodb_page_size or instant ADD/DROP COLUMN. They may be deprecated in the future. We do not want an unnecessary file format change for them. The new full_crc32() format also cleans up the MariaDB tablespace flags. We will reserve flags to store the page_compressed compression algorithm, and to store the compressed payload length, so that checksum can be computed over the compressed (and possibly encrypted) stream and can be validated without decrypting or decompressing the page. In the full_crc32 format, there no longer are separate before-encryption and after-encryption checksums for pages. The single checksum is computed on the page contents that is written to the file. We do not make the new algorithm the default for two reasons. First, MariaDB 10.4.2 was a beta release, and the default values of parameters should not change after beta. Second, we did not yet implement the full_crc32 format for page_compressed pages. This will be fixed in MDEV-18644. This is joint work with Marko Mäkelä.
2019-02-19 20:00:00 +01:00
/** Assign the full crc32 checksum for non-compressed page.
@param[in,out] page page to be updated */
void buf_flush_assign_full_crc32_checksum(byte* page)
{
MDEV-18644: Support full_crc32 for page_compressed This is a follow-up task to MDEV-12026, which introduced innodb_checksum_algorithm=full_crc32 and a simpler page format. MDEV-12026 did not enable full_crc32 for page_compressed tables, which we will be doing now. This is joint work with Thirunarayanan Balathandayuthapani. For innodb_checksum_algorithm=full_crc32 we change the page_compressed format as follows: FIL_PAGE_TYPE: The most significant bit will be set to indicate page_compressed format. The least significant bits will contain the compressed page size, rounded up to a multiple of 256 bytes. The checksum will be stored in the last 4 bytes of the page (whether it is the full page or a page_compressed page whose size is determined by FIL_PAGE_TYPE), covering all preceding bytes of the page. If encryption is used, then the page will be encrypted between compression and computing the checksum. For page_compressed, FIL_PAGE_LSN will not be repeated at the end of the page. FSP_SPACE_FLAGS (already implemented as part of MDEV-12026): We will store the innodb_compression_algorithm that may be used to compress pages. Previously, the choice of algorithm was written to each compressed data page separately, and one would be unable to know in advance which compression algorithm(s) are used. fil_space_t::full_crc32_page_compressed_len(): Determine if the page_compressed algorithm of the tablespace needs to know the exact length of the compressed data. If yes, we will reserve and write an extra byte for this right before the checksum. buf_page_is_compressed(): Determine if a page uses page_compressed (in any innodb_checksum_algorithm). fil_page_decompress(): Pass also fil_space_t::flags so that the format can be determined. buf_page_is_zeroes(): Check if a page is full of zero bytes. buf_page_full_crc32_is_corrupted(): Renamed from buf_encrypted_full_crc32_page_is_corrupted(). For full_crc32, we always simply validate the checksum to the page contents, while the physical page size is explicitly specified by an unencrypted part of the page header. buf_page_full_crc32_size(): Determine the size of a full_crc32 page. buf_dblwr_check_page_lsn(): Make this a debug-only function, because it involves potentially costly lookups of fil_space_t. create_table_info_t::check_table_options(), ha_innobase::check_if_supported_inplace_alter(): Do allow the creation of SPATIAL INDEX with full_crc32 also when page_compressed is used. commit_cache_norebuild(): Preserve the compression algorithm when updating the page_compression_level. dict_tf_to_fsp_flags(): Set the flags for page compression algorithm. FIXME: Maybe there should be a table option page_compression_algorithm and a session variable to back it?
2019-03-18 13:08:43 +01:00
ut_d(bool compressed = false);
ut_d(bool corrupted = false);
ut_d(const uint size = buf_page_full_crc32_size(page, &compressed,
&corrupted));
ut_ad(!compressed);
ut_ad(!corrupted);
ut_ad(size == uint(srv_page_size));
const ulint payload = srv_page_size - FIL_PAGE_FCRC32_CHECKSUM;
mach_write_to_4(page + payload, ut_crc32(page, payload));
MDEV-12026: Implement innodb_checksum_algorithm=full_crc32 MariaDB data-at-rest encryption (innodb_encrypt_tables) had repurposed the same unused data field that was repurposed in MySQL 5.7 (and MariaDB 10.2) for the Split Sequence Number (SSN) field of SPATIAL INDEX. Because of this, MariaDB was unable to support encryption on SPATIAL INDEX pages. Furthermore, InnoDB page checksums skipped some bytes, and there are multiple variations and checksum algorithms. By default, InnoDB accepts all variations of all algorithms that ever existed. This unnecessarily weakens the page checksums. We hereby introduce two more innodb_checksum_algorithm variants (full_crc32, strict_full_crc32) that are special in a way: When either setting is active, newly created data files will carry a flag (fil_space_t::full_crc32()) that indicates that all pages of the file will use a full CRC-32C checksum over the entire page contents (excluding the bytes where the checksum is stored, at the very end of the page). Such files will always use that checksum, no matter what the parameter innodb_checksum_algorithm is assigned to. For old files, the old checksum algorithms will continue to be used. The value strict_full_crc32 will be equivalent to strict_crc32 and the value full_crc32 will be equivalent to crc32. ROW_FORMAT=COMPRESSED tables will only use the old format. These tables do not support new features, such as larger innodb_page_size or instant ADD/DROP COLUMN. They may be deprecated in the future. We do not want an unnecessary file format change for them. The new full_crc32() format also cleans up the MariaDB tablespace flags. We will reserve flags to store the page_compressed compression algorithm, and to store the compressed payload length, so that checksum can be computed over the compressed (and possibly encrypted) stream and can be validated without decrypting or decompressing the page. In the full_crc32 format, there no longer are separate before-encryption and after-encryption checksums for pages. The single checksum is computed on the page contents that is written to the file. We do not make the new algorithm the default for two reasons. First, MariaDB 10.4.2 was a beta release, and the default values of parameters should not change after beta. Second, we did not yet implement the full_crc32 format for page_compressed pages. This will be fixed in MDEV-18644. This is joint work with Marko Mäkelä.
2019-02-19 20:00:00 +01:00
}
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
/** Initialize a page for writing to the tablespace.
MDEV-12026: Implement innodb_checksum_algorithm=full_crc32 MariaDB data-at-rest encryption (innodb_encrypt_tables) had repurposed the same unused data field that was repurposed in MySQL 5.7 (and MariaDB 10.2) for the Split Sequence Number (SSN) field of SPATIAL INDEX. Because of this, MariaDB was unable to support encryption on SPATIAL INDEX pages. Furthermore, InnoDB page checksums skipped some bytes, and there are multiple variations and checksum algorithms. By default, InnoDB accepts all variations of all algorithms that ever existed. This unnecessarily weakens the page checksums. We hereby introduce two more innodb_checksum_algorithm variants (full_crc32, strict_full_crc32) that are special in a way: When either setting is active, newly created data files will carry a flag (fil_space_t::full_crc32()) that indicates that all pages of the file will use a full CRC-32C checksum over the entire page contents (excluding the bytes where the checksum is stored, at the very end of the page). Such files will always use that checksum, no matter what the parameter innodb_checksum_algorithm is assigned to. For old files, the old checksum algorithms will continue to be used. The value strict_full_crc32 will be equivalent to strict_crc32 and the value full_crc32 will be equivalent to crc32. ROW_FORMAT=COMPRESSED tables will only use the old format. These tables do not support new features, such as larger innodb_page_size or instant ADD/DROP COLUMN. They may be deprecated in the future. We do not want an unnecessary file format change for them. The new full_crc32() format also cleans up the MariaDB tablespace flags. We will reserve flags to store the page_compressed compression algorithm, and to store the compressed payload length, so that checksum can be computed over the compressed (and possibly encrypted) stream and can be validated without decrypting or decompressing the page. In the full_crc32 format, there no longer are separate before-encryption and after-encryption checksums for pages. The single checksum is computed on the page contents that is written to the file. We do not make the new algorithm the default for two reasons. First, MariaDB 10.4.2 was a beta release, and the default values of parameters should not change after beta. Second, we did not yet implement the full_crc32 format for page_compressed pages. This will be fixed in MDEV-18644. This is joint work with Marko Mäkelä.
2019-02-19 20:00:00 +01:00
@param[in] block buffer block; NULL if bypassing
the buffer pool
@param[in,out] page page frame
@param[in,out] page_zip_ compressed page, or NULL if
uncompressed
@param[in] use_full_checksum whether tablespace uses full checksum */
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
void
buf_flush_init_for_writing(
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
const buf_block_t* block,
byte* page,
void* page_zip_,
MDEV-12026: Implement innodb_checksum_algorithm=full_crc32 MariaDB data-at-rest encryption (innodb_encrypt_tables) had repurposed the same unused data field that was repurposed in MySQL 5.7 (and MariaDB 10.2) for the Split Sequence Number (SSN) field of SPATIAL INDEX. Because of this, MariaDB was unable to support encryption on SPATIAL INDEX pages. Furthermore, InnoDB page checksums skipped some bytes, and there are multiple variations and checksum algorithms. By default, InnoDB accepts all variations of all algorithms that ever existed. This unnecessarily weakens the page checksums. We hereby introduce two more innodb_checksum_algorithm variants (full_crc32, strict_full_crc32) that are special in a way: When either setting is active, newly created data files will carry a flag (fil_space_t::full_crc32()) that indicates that all pages of the file will use a full CRC-32C checksum over the entire page contents (excluding the bytes where the checksum is stored, at the very end of the page). Such files will always use that checksum, no matter what the parameter innodb_checksum_algorithm is assigned to. For old files, the old checksum algorithms will continue to be used. The value strict_full_crc32 will be equivalent to strict_crc32 and the value full_crc32 will be equivalent to crc32. ROW_FORMAT=COMPRESSED tables will only use the old format. These tables do not support new features, such as larger innodb_page_size or instant ADD/DROP COLUMN. They may be deprecated in the future. We do not want an unnecessary file format change for them. The new full_crc32() format also cleans up the MariaDB tablespace flags. We will reserve flags to store the page_compressed compression algorithm, and to store the compressed payload length, so that checksum can be computed over the compressed (and possibly encrypted) stream and can be validated without decrypting or decompressing the page. In the full_crc32 format, there no longer are separate before-encryption and after-encryption checksums for pages. The single checksum is computed on the page contents that is written to the file. We do not make the new algorithm the default for two reasons. First, MariaDB 10.4.2 was a beta release, and the default values of parameters should not change after beta. Second, we did not yet implement the full_crc32 format for page_compressed pages. This will be fixed in MDEV-18644. This is joint work with Marko Mäkelä.
2019-02-19 20:00:00 +01:00
bool use_full_checksum)
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
{
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
if (block && block->page.frame != page) {
MDEV-12026: Implement innodb_checksum_algorithm=full_crc32 MariaDB data-at-rest encryption (innodb_encrypt_tables) had repurposed the same unused data field that was repurposed in MySQL 5.7 (and MariaDB 10.2) for the Split Sequence Number (SSN) field of SPATIAL INDEX. Because of this, MariaDB was unable to support encryption on SPATIAL INDEX pages. Furthermore, InnoDB page checksums skipped some bytes, and there are multiple variations and checksum algorithms. By default, InnoDB accepts all variations of all algorithms that ever existed. This unnecessarily weakens the page checksums. We hereby introduce two more innodb_checksum_algorithm variants (full_crc32, strict_full_crc32) that are special in a way: When either setting is active, newly created data files will carry a flag (fil_space_t::full_crc32()) that indicates that all pages of the file will use a full CRC-32C checksum over the entire page contents (excluding the bytes where the checksum is stored, at the very end of the page). Such files will always use that checksum, no matter what the parameter innodb_checksum_algorithm is assigned to. For old files, the old checksum algorithms will continue to be used. The value strict_full_crc32 will be equivalent to strict_crc32 and the value full_crc32 will be equivalent to crc32. ROW_FORMAT=COMPRESSED tables will only use the old format. These tables do not support new features, such as larger innodb_page_size or instant ADD/DROP COLUMN. They may be deprecated in the future. We do not want an unnecessary file format change for them. The new full_crc32() format also cleans up the MariaDB tablespace flags. We will reserve flags to store the page_compressed compression algorithm, and to store the compressed payload length, so that checksum can be computed over the compressed (and possibly encrypted) stream and can be validated without decrypting or decompressing the page. In the full_crc32 format, there no longer are separate before-encryption and after-encryption checksums for pages. The single checksum is computed on the page contents that is written to the file. We do not make the new algorithm the default for two reasons. First, MariaDB 10.4.2 was a beta release, and the default values of parameters should not change after beta. Second, we did not yet implement the full_crc32 format for page_compressed pages. This will be fixed in MDEV-18644. This is joint work with Marko Mäkelä.
2019-02-19 20:00:00 +01:00
/* If page is encrypted in full crc32 format then
checksum stored already as a part of fil_encrypt_buf() */
ut_ad(use_full_checksum);
return;
}
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
ut_ad(!block || block->page.frame == page);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
ut_ad(page);
if (page_zip_) {
page_zip_des_t* page_zip;
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
ulint size;
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
page_zip = static_cast<page_zip_des_t*>(page_zip_);
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
ut_ad(!block || &block->page.zip == page_zip);
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
size = page_zip_get_size(page_zip);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
ut_ad(size);
ut_ad(ut_is_2pow(size));
ut_ad(size <= UNIV_ZIP_SIZE_MAX);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
switch (fil_page_get_type(page)) {
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
case FIL_PAGE_TYPE_ALLOCATED:
case FIL_PAGE_INODE:
case FIL_PAGE_IBUF_BITMAP:
case FIL_PAGE_TYPE_FSP_HDR:
case FIL_PAGE_TYPE_XDES:
/* These are essentially uncompressed pages. */
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
memcpy(page_zip->data, page, size);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
/* fall through */
case FIL_PAGE_TYPE_ZBLOB:
case FIL_PAGE_TYPE_ZBLOB2:
case FIL_PAGE_INDEX:
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
case FIL_PAGE_RTREE:
MDEV-21132 Remove buf_page_t::newest_modification At each mini-transaction commit, the log sequence number of the mini-transaction must be written to each modified page, so that it will be available in the FIL_PAGE_LSN field when the page is being read in crash recovery. InnoDB was unnecessarily allocating redundant storage for the field, in buf_page_t::newest_modification. Let us access FIL_PAGE_LSN directly. Furthermore, on ALTER TABLE...IMPORT TABLESPACE, let us write 0 to FIL_PAGE_LSN instead of using log_sys.lsn. buf_flush_init_for_writing(), buf_flush_update_zip_checksum(), fil_encrypt_buf_for_full_crc32(), fil_encrypt_buf(), fil_space_encrypt(): Remove the parameter lsn. buf_page_get_newest_modification(): Merge with the only caller. buf_tmp_reserve_compression_buf(), buf_tmp_page_encrypt(), buf_page_encrypt(): Define static in the same compilation unit with the only caller. PageConverter::m_current_lsn: Remove. Write 0 to FIL_PAGE_LSN on ALTER TABLE...IMPORT TABLESPACE.
2019-11-25 08:39:51 +01:00
buf_flush_update_zip_checksum(page_zip->data, size);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
return;
}
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
ib::error() << "The compressed page to be written"
" seems corrupt:";
ut_print_buf(stderr, page, size);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
fputs("\nInnoDB: Possibly older version of the page:", stderr);
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
ut_print_buf(stderr, page_zip->data, size);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
putc('\n', stderr);
ut_error;
}
MDEV-12026: Implement innodb_checksum_algorithm=full_crc32 MariaDB data-at-rest encryption (innodb_encrypt_tables) had repurposed the same unused data field that was repurposed in MySQL 5.7 (and MariaDB 10.2) for the Split Sequence Number (SSN) field of SPATIAL INDEX. Because of this, MariaDB was unable to support encryption on SPATIAL INDEX pages. Furthermore, InnoDB page checksums skipped some bytes, and there are multiple variations and checksum algorithms. By default, InnoDB accepts all variations of all algorithms that ever existed. This unnecessarily weakens the page checksums. We hereby introduce two more innodb_checksum_algorithm variants (full_crc32, strict_full_crc32) that are special in a way: When either setting is active, newly created data files will carry a flag (fil_space_t::full_crc32()) that indicates that all pages of the file will use a full CRC-32C checksum over the entire page contents (excluding the bytes where the checksum is stored, at the very end of the page). Such files will always use that checksum, no matter what the parameter innodb_checksum_algorithm is assigned to. For old files, the old checksum algorithms will continue to be used. The value strict_full_crc32 will be equivalent to strict_crc32 and the value full_crc32 will be equivalent to crc32. ROW_FORMAT=COMPRESSED tables will only use the old format. These tables do not support new features, such as larger innodb_page_size or instant ADD/DROP COLUMN. They may be deprecated in the future. We do not want an unnecessary file format change for them. The new full_crc32() format also cleans up the MariaDB tablespace flags. We will reserve flags to store the page_compressed compression algorithm, and to store the compressed payload length, so that checksum can be computed over the compressed (and possibly encrypted) stream and can be validated without decrypting or decompressing the page. In the full_crc32 format, there no longer are separate before-encryption and after-encryption checksums for pages. The single checksum is computed on the page contents that is written to the file. We do not make the new algorithm the default for two reasons. First, MariaDB 10.4.2 was a beta release, and the default values of parameters should not change after beta. Second, we did not yet implement the full_crc32 format for page_compressed pages. This will be fixed in MDEV-18644. This is joint work with Marko Mäkelä.
2019-02-19 20:00:00 +01:00
if (use_full_checksum) {
MDEV-21133 Optimize access to InnoDB page header fields Introduce memcpy_aligned<N>(), memcmp_aligned<N>(), memset_aligned<N>() and use them for accessing InnoDB page header fields that are known to be aligned. MY_ASSUME_ALIGNED(): Wrapper for the GCC/clang __builtin_assume_aligned(). Nothing similar seems to exist in Microsoft Visual Studio, and the C++20 std::assume_aligned is not available to us yet. Explicitly specified alignment guarantees allow compilers to generate faster code on platforms with strict alignment rules, instead of emitting calls to potentially unaligned memcpy(), memcmp(), or memset().
2019-11-26 09:14:07 +01:00
static_assert(FIL_PAGE_FCRC32_END_LSN % 4 == 0, "aligned");
static_assert(FIL_PAGE_LSN % 4 == 0, "aligned");
memcpy_aligned<4>(page + srv_page_size
- FIL_PAGE_FCRC32_END_LSN,
FIL_PAGE_LSN + 4 + page, 4);
Cleanup after MDEV-12026 buf_flush_init_for_writing(): If innodb_checksum_algorithm=full_crc32, return even earlier. The logic to reset garbage FIL_PAGE_TYPE fields applies to data pages in old data files that were last written before commit 3926673ce7149aa223103126b6aeac819b10fab5 was introduced. When innodb_checksum_algorithm=full_crc32 is in effect, it only applies to newly created data files. A file that was created in full_crc32 format will keep that format. Other files will keep some other format than full_crc32. Hence, the garbage-resetting logic does not apply full_crc32 files.
2019-12-30 17:59:02 +01:00
return buf_flush_assign_full_crc32_checksum(page);
MDEV-12026: Implement innodb_checksum_algorithm=full_crc32 MariaDB data-at-rest encryption (innodb_encrypt_tables) had repurposed the same unused data field that was repurposed in MySQL 5.7 (and MariaDB 10.2) for the Split Sequence Number (SSN) field of SPATIAL INDEX. Because of this, MariaDB was unable to support encryption on SPATIAL INDEX pages. Furthermore, InnoDB page checksums skipped some bytes, and there are multiple variations and checksum algorithms. By default, InnoDB accepts all variations of all algorithms that ever existed. This unnecessarily weakens the page checksums. We hereby introduce two more innodb_checksum_algorithm variants (full_crc32, strict_full_crc32) that are special in a way: When either setting is active, newly created data files will carry a flag (fil_space_t::full_crc32()) that indicates that all pages of the file will use a full CRC-32C checksum over the entire page contents (excluding the bytes where the checksum is stored, at the very end of the page). Such files will always use that checksum, no matter what the parameter innodb_checksum_algorithm is assigned to. For old files, the old checksum algorithms will continue to be used. The value strict_full_crc32 will be equivalent to strict_crc32 and the value full_crc32 will be equivalent to crc32. ROW_FORMAT=COMPRESSED tables will only use the old format. These tables do not support new features, such as larger innodb_page_size or instant ADD/DROP COLUMN. They may be deprecated in the future. We do not want an unnecessary file format change for them. The new full_crc32() format also cleans up the MariaDB tablespace flags. We will reserve flags to store the page_compressed compression algorithm, and to store the compressed payload length, so that checksum can be computed over the compressed (and possibly encrypted) stream and can be validated without decrypting or decompressing the page. In the full_crc32 format, there no longer are separate before-encryption and after-encryption checksums for pages. The single checksum is computed on the page contents that is written to the file. We do not make the new algorithm the default for two reasons. First, MariaDB 10.4.2 was a beta release, and the default values of parameters should not change after beta. Second, we did not yet implement the full_crc32 format for page_compressed pages. This will be fixed in MDEV-18644. This is joint work with Marko Mäkelä.
2019-02-19 20:00:00 +01:00
}
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
Merge 10.4 into 10.5
2020-01-03 15:15:40 +01:00
static_assert(FIL_PAGE_END_LSN_OLD_CHKSUM % 8 == 0, "aligned");
static_assert(FIL_PAGE_LSN % 8 == 0, "aligned");
memcpy_aligned<8>(page + srv_page_size - FIL_PAGE_END_LSN_OLD_CHKSUM,
FIL_PAGE_LSN + page, 8);
MDEV-15524 Do not disable page checksums for temporary tables buf_flush_init_for_writing(): Remove the parameter skip_checksum.
2018-03-09 12:43:32 +01:00
if (block && srv_page_size == 16384) {
/* The page type could be garbage in old files
created before MySQL 5.5. Such files always
had a page size of 16 kilobytes. */
ulint page_type = fil_page_get_type(page);
ulint reset_type = page_type;
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
switch (block->page.id().page_no() % 16384) {
MDEV-15524 Do not disable page checksums for temporary tables buf_flush_init_for_writing(): Remove the parameter skip_checksum.
2018-03-09 12:43:32 +01:00
case 0:
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
reset_type = block->page.id().page_no() == 0
MDEV-15524 Do not disable page checksums for temporary tables buf_flush_init_for_writing(): Remove the parameter skip_checksum.
2018-03-09 12:43:32 +01:00
? FIL_PAGE_TYPE_FSP_HDR
: FIL_PAGE_TYPE_XDES;
break;
case 1:
reset_type = FIL_PAGE_IBUF_BITMAP;
break;
case FSP_TRX_SYS_PAGE_NO:
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
if (block->page.id()
== page_id_t(TRX_SYS_SPACE, TRX_SYS_PAGE_NO)) {
MDEV-15524 Do not disable page checksums for temporary tables buf_flush_init_for_writing(): Remove the parameter skip_checksum.
2018-03-09 12:43:32 +01:00
reset_type = FIL_PAGE_TYPE_TRX_SYS;
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
break;
MDEV-15524 Do not disable page checksums for temporary tables buf_flush_init_for_writing(): Remove the parameter skip_checksum.
2018-03-09 12:43:32 +01:00
}
/* fall through */
default:
switch (page_type) {
case FIL_PAGE_INDEX:
Merge bb-10.2-ext into 10.3
2018-03-13 07:15:06 +01:00
case FIL_PAGE_TYPE_INSTANT:
MDEV-15524 Do not disable page checksums for temporary tables buf_flush_init_for_writing(): Remove the parameter skip_checksum.
2018-03-09 12:43:32 +01:00
case FIL_PAGE_RTREE:
case FIL_PAGE_UNDO_LOG:
case FIL_PAGE_INODE:
case FIL_PAGE_IBUF_FREE_LIST:
case FIL_PAGE_TYPE_ALLOCATED:
case FIL_PAGE_TYPE_SYS:
case FIL_PAGE_TYPE_TRX_SYS:
case FIL_PAGE_TYPE_BLOB:
case FIL_PAGE_TYPE_ZBLOB:
case FIL_PAGE_TYPE_ZBLOB2:
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
break;
MDEV-15524 Do not disable page checksums for temporary tables buf_flush_init_for_writing(): Remove the parameter skip_checksum.
2018-03-09 12:43:32 +01:00
case FIL_PAGE_TYPE_FSP_HDR:
case FIL_PAGE_TYPE_XDES:
case FIL_PAGE_IBUF_BITMAP:
/* These pages should have
predetermined page numbers
(see above). */
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
default:
MDEV-15524 Do not disable page checksums for temporary tables buf_flush_init_for_writing(): Remove the parameter skip_checksum.
2018-03-09 12:43:32 +01:00
reset_type = FIL_PAGE_TYPE_UNKNOWN;
break;
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
}
}
MDEV-15524 Do not disable page checksums for temporary tables buf_flush_init_for_writing(): Remove the parameter skip_checksum.
2018-03-09 12:43:32 +01:00
if (UNIV_UNLIKELY(page_type != reset_type)) {
ib::info()
<< "Resetting invalid page "
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
<< block->page.id() << " type "
MDEV-15524 Do not disable page checksums for temporary tables buf_flush_init_for_writing(): Remove the parameter skip_checksum.
2018-03-09 12:43:32 +01:00
<< page_type << " to "
<< reset_type << " when flushing.";
fil_page_set_type(page, reset_type);
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
}
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
}
MDEV-25105 Remove innodb_checksum_algorithm values none,innodb,... Historically, InnoDB supported a buggy page checksum algorithm that did not compute a checksum over the full page. Later, well before MySQL 4.1 introduced .ibd files and the innodb_file_per_table option, the algorithm was corrected and the first 4 bytes of each page were redefined to be a checksum. The original checksum was so slow that an option to disable page checksum was introduced for benchmarketing purposes. The Intel Nehalem microarchitecture introduced the SSE4.2 instruction set extension, which includes instructions for faster computation of CRC-32C. In MySQL 5.6 (and MariaDB 10.0), innodb_checksum_algorithm=crc32 was implemented to make of that. As that option was changed to be the default in MySQL 5.7, a bug was found on big-endian platforms and some work-around code was added to weaken that checksum further. MariaDB disables that work-around by default since MDEV-17958. Later, SIMD-accelerated CRC-32C has been implemented in MariaDB for POWER and ARM and also for IA-32/AMD64, making use of carry-less multiplication where available. Long story short, innodb_checksum_algorithm=crc32 is faster and more secure than the pre-MySQL 5.6 checksum, called innodb_checksum_algorithm=innodb. It should have removed any need to use innodb_checksum_algorithm=none. The setting innodb_checksum_algorithm=crc32 is the default in MySQL 5.7 and MariaDB Server 10.2, 10.3, 10.4. In MariaDB 10.5, MDEV-19534 made innodb_checksum_algorithm=full_crc32 the default. It is even faster and more secure. The default settings in MariaDB do allow old data files to be read, no matter if a worse checksum algorithm had been used. (Unfortunately, before innodb_checksum_algorithm=full_crc32, the data files did not identify which checksum algorithm is being used.) The non-default settings innodb_checksum_algorithm=strict_crc32 or innodb_checksum_algorithm=strict_full_crc32 would only allow CRC-32C checksums. The incompatibility with old data files is why they are not the default. The newest server not to support innodb_checksum_algorithm=crc32 were MySQL 5.5 and MariaDB 5.5. Both have reached their end of life. A valid reason for using innodb_checksum_algorithm=innodb could have been the ability to downgrade. If it is really needed, data files can be converted with an older version of the innochecksum utility. Because there is no good reason to allow data files to be written with insecure checksums, we will reject those option values: innodb_checksum_algorithm=none innodb_checksum_algorithm=innodb innodb_checksum_algorithm=strict_none innodb_checksum_algorithm=strict_innodb Furthermore, the following innochecksum options will be removed, because only strict crc32 will be supported: innochecksum --strict-check=crc32 innochecksum -C crc32 innochecksum --write=crc32 innochecksum -w crc32 If a user wishes to convert a data file to use a different checksum (so that it might be used with the no-longer-supported MySQL 5.5 or MariaDB 5.5, which do not support IMPORT TABLESPACE nor system tablespace format changes that were made in MariaDB 10.3), then the innochecksum tool from MariaDB 10.2, 10.3, 10.4, 10.5 or MySQL 5.7 can be used. Reviewed by: Thirunarayanan Balathandayuthapani
2021-03-11 10:56:35 +01:00
const uint32_t checksum = buf_calc_page_crc32(page);
mach_write_to_4(page + FIL_PAGE_SPACE_OR_CHKSUM, checksum);
Replace univ_page_size and UNIV_PAGE_SIZE Try to use one variable (srv_page_size) for innodb_page_size. Also, replace UNIV_PAGE_SIZE_SHIFT with srv_page_size_shift.
2018-04-27 12:49:25 +02:00
mach_write_to_4(page + srv_page_size - FIL_PAGE_END_LSN_OLD_CHKSUM,
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
checksum);
}
MDEV-21132 Remove buf_page_t::newest_modification At each mini-transaction commit, the log sequence number of the mini-transaction must be written to each modified page, so that it will be available in the FIL_PAGE_LSN field when the page is being read in crash recovery. InnoDB was unnecessarily allocating redundant storage for the field, in buf_page_t::newest_modification. Let us access FIL_PAGE_LSN directly. Furthermore, on ALTER TABLE...IMPORT TABLESPACE, let us write 0 to FIL_PAGE_LSN instead of using log_sys.lsn. buf_flush_init_for_writing(), buf_flush_update_zip_checksum(), fil_encrypt_buf_for_full_crc32(), fil_encrypt_buf(), fil_space_encrypt(): Remove the parameter lsn. buf_page_get_newest_modification(): Merge with the only caller. buf_tmp_reserve_compression_buf(), buf_tmp_page_encrypt(), buf_page_encrypt(): Define static in the same compilation unit with the only caller. PageConverter::m_current_lsn: Remove. Write 0 to FIL_PAGE_LSN on ALTER TABLE...IMPORT TABLESPACE.
2019-11-25 08:39:51 +01:00
/** Reserve a buffer for compression.
@param[in,out] slot reserved slot */
static void buf_tmp_reserve_compression_buf(buf_tmp_buffer_t* slot)
{
if (slot->comp_buf)
return;
/* Both Snappy and LZO compression methods require that the output
buffer be bigger than input buffer. Adjust the allocated size. */
ulint size= srv_page_size;
#ifdef HAVE_LZO
size+= LZO1X_1_15_MEM_COMPRESS;
#elif defined HAVE_SNAPPY
size= snappy_max_compressed_length(size);
#endif
slot->comp_buf= static_cast<byte*>(aligned_malloc(size, srv_page_size));
}
/** Encrypt a buffer of temporary tablespace
@param[in] offset Page offset
@param[in] s Page to encrypt
@param[in,out] d Output buffer
@return encrypted buffer or NULL */
static byte* buf_tmp_page_encrypt(ulint offset, const byte* s, byte* d)
{
/* Calculate the start offset in a page */
MDEV-21907: Fix most clang -Wconversion in InnoDB Declare innodb_purge_threads as 4-byte integer (UINT) instead of 4-or-8-byte (ULONG) and adjust the documentation string.
2020-03-10 19:05:17 +01:00
uint srclen= static_cast<uint>(srv_page_size) -
(FIL_PAGE_FILE_FLUSH_LSN_OR_KEY_VERSION +
FIL_PAGE_FCRC32_CHECKSUM);
MDEV-21132 Remove buf_page_t::newest_modification At each mini-transaction commit, the log sequence number of the mini-transaction must be written to each modified page, so that it will be available in the FIL_PAGE_LSN field when the page is being read in crash recovery. InnoDB was unnecessarily allocating redundant storage for the field, in buf_page_t::newest_modification. Let us access FIL_PAGE_LSN directly. Furthermore, on ALTER TABLE...IMPORT TABLESPACE, let us write 0 to FIL_PAGE_LSN instead of using log_sys.lsn. buf_flush_init_for_writing(), buf_flush_update_zip_checksum(), fil_encrypt_buf_for_full_crc32(), fil_encrypt_buf(), fil_space_encrypt(): Remove the parameter lsn. buf_page_get_newest_modification(): Merge with the only caller. buf_tmp_reserve_compression_buf(), buf_tmp_page_encrypt(), buf_page_encrypt(): Define static in the same compilation unit with the only caller. PageConverter::m_current_lsn: Remove. Write 0 to FIL_PAGE_LSN on ALTER TABLE...IMPORT TABLESPACE.
2019-11-25 08:39:51 +01:00
const byte* src= s + FIL_PAGE_FILE_FLUSH_LSN_OR_KEY_VERSION;
byte* dst= d + FIL_PAGE_FILE_FLUSH_LSN_OR_KEY_VERSION;
memcpy(d, s, FIL_PAGE_FILE_FLUSH_LSN_OR_KEY_VERSION);
if (!log_tmp_block_encrypt(src, srclen, dst, (offset * srv_page_size), true))
return NULL;
const ulint payload= srv_page_size - FIL_PAGE_FCRC32_CHECKSUM;
mach_write_to_4(d + payload, ut_crc32(d, payload));
srv_stats.pages_encrypted.inc();
srv_stats.n_temp_blocks_encrypted.inc();
return d;
}
/** Encryption and page_compression hook that is called just before
a page is written to disk.
@param[in,out] space tablespace
@param[in,out] bpage buffer page
@param[in] s physical page frame that is being encrypted
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
@param[in,out] size payload size in bytes
MDEV-21132 Remove buf_page_t::newest_modification At each mini-transaction commit, the log sequence number of the mini-transaction must be written to each modified page, so that it will be available in the FIL_PAGE_LSN field when the page is being read in crash recovery. InnoDB was unnecessarily allocating redundant storage for the field, in buf_page_t::newest_modification. Let us access FIL_PAGE_LSN directly. Furthermore, on ALTER TABLE...IMPORT TABLESPACE, let us write 0 to FIL_PAGE_LSN instead of using log_sys.lsn. buf_flush_init_for_writing(), buf_flush_update_zip_checksum(), fil_encrypt_buf_for_full_crc32(), fil_encrypt_buf(), fil_space_encrypt(): Remove the parameter lsn. buf_page_get_newest_modification(): Merge with the only caller. buf_tmp_reserve_compression_buf(), buf_tmp_page_encrypt(), buf_page_encrypt(): Define static in the same compilation unit with the only caller. PageConverter::m_current_lsn: Remove. Write 0 to FIL_PAGE_LSN on ALTER TABLE...IMPORT TABLESPACE.
2019-11-25 08:39:51 +01:00
@return page frame to be written to file
(may be src_frame or an encrypted/compressed copy of it) */
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
static byte *buf_page_encrypt(fil_space_t* space, buf_page_t* bpage, byte* s,
MDEV-27058: Move buf_page_t::slot to IORequest::slot MDEV-23855 and MDEV-23399 already moved some transient data fields from buffer pool page descriptors to IORequest, but the write buffer of PAGE_COMPRESSED or ENCRYPTED tables was missed. Since is only needed during asynchronous page write requests, it belongs to IORequest.
2021-11-16 17:16:53 +01:00
buf_tmp_buffer_t **slot, size_t *size)
MDEV-21132 Remove buf_page_t::newest_modification At each mini-transaction commit, the log sequence number of the mini-transaction must be written to each modified page, so that it will be available in the FIL_PAGE_LSN field when the page is being read in crash recovery. InnoDB was unnecessarily allocating redundant storage for the field, in buf_page_t::newest_modification. Let us access FIL_PAGE_LSN directly. Furthermore, on ALTER TABLE...IMPORT TABLESPACE, let us write 0 to FIL_PAGE_LSN instead of using log_sys.lsn. buf_flush_init_for_writing(), buf_flush_update_zip_checksum(), fil_encrypt_buf_for_full_crc32(), fil_encrypt_buf(), fil_space_encrypt(): Remove the parameter lsn. buf_page_get_newest_modification(): Merge with the only caller. buf_tmp_reserve_compression_buf(), buf_tmp_page_encrypt(), buf_page_encrypt(): Define static in the same compilation unit with the only caller. PageConverter::m_current_lsn: Remove. Write 0 to FIL_PAGE_LSN on ALTER TABLE...IMPORT TABLESPACE.
2019-11-25 08:39:51 +01:00
{
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
ut_ad(!bpage->is_freed());
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
ut_ad(space->id == bpage->id().space());
MDEV-27058: Move buf_page_t::slot to IORequest::slot MDEV-23855 and MDEV-23399 already moved some transient data fields from buffer pool page descriptors to IORequest, but the write buffer of PAGE_COMPRESSED or ENCRYPTED tables was missed. Since is only needed during asynchronous page write requests, it belongs to IORequest.
2021-11-16 17:16:53 +01:00
ut_ad(!*slot);
MDEV-21132 Remove buf_page_t::newest_modification At each mini-transaction commit, the log sequence number of the mini-transaction must be written to each modified page, so that it will be available in the FIL_PAGE_LSN field when the page is being read in crash recovery. InnoDB was unnecessarily allocating redundant storage for the field, in buf_page_t::newest_modification. Let us access FIL_PAGE_LSN directly. Furthermore, on ALTER TABLE...IMPORT TABLESPACE, let us write 0 to FIL_PAGE_LSN instead of using log_sys.lsn. buf_flush_init_for_writing(), buf_flush_update_zip_checksum(), fil_encrypt_buf_for_full_crc32(), fil_encrypt_buf(), fil_space_encrypt(): Remove the parameter lsn. buf_page_get_newest_modification(): Merge with the only caller. buf_tmp_reserve_compression_buf(), buf_tmp_page_encrypt(), buf_page_encrypt(): Define static in the same compilation unit with the only caller. PageConverter::m_current_lsn: Remove. Write 0 to FIL_PAGE_LSN on ALTER TABLE...IMPORT TABLESPACE.
2019-11-25 08:39:51 +01:00
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
const uint32_t page_no= bpage->id().page_no();
MDEV-21132 Remove buf_page_t::newest_modification At each mini-transaction commit, the log sequence number of the mini-transaction must be written to each modified page, so that it will be available in the FIL_PAGE_LSN field when the page is being read in crash recovery. InnoDB was unnecessarily allocating redundant storage for the field, in buf_page_t::newest_modification. Let us access FIL_PAGE_LSN directly. Furthermore, on ALTER TABLE...IMPORT TABLESPACE, let us write 0 to FIL_PAGE_LSN instead of using log_sys.lsn. buf_flush_init_for_writing(), buf_flush_update_zip_checksum(), fil_encrypt_buf_for_full_crc32(), fil_encrypt_buf(), fil_space_encrypt(): Remove the parameter lsn. buf_page_get_newest_modification(): Merge with the only caller. buf_tmp_reserve_compression_buf(), buf_tmp_page_encrypt(), buf_page_encrypt(): Define static in the same compilation unit with the only caller. PageConverter::m_current_lsn: Remove. Write 0 to FIL_PAGE_LSN on ALTER TABLE...IMPORT TABLESPACE.
2019-11-25 08:39:51 +01:00
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
switch (page_no) {
MDEV-21132 Remove buf_page_t::newest_modification At each mini-transaction commit, the log sequence number of the mini-transaction must be written to each modified page, so that it will be available in the FIL_PAGE_LSN field when the page is being read in crash recovery. InnoDB was unnecessarily allocating redundant storage for the field, in buf_page_t::newest_modification. Let us access FIL_PAGE_LSN directly. Furthermore, on ALTER TABLE...IMPORT TABLESPACE, let us write 0 to FIL_PAGE_LSN instead of using log_sys.lsn. buf_flush_init_for_writing(), buf_flush_update_zip_checksum(), fil_encrypt_buf_for_full_crc32(), fil_encrypt_buf(), fil_space_encrypt(): Remove the parameter lsn. buf_page_get_newest_modification(): Merge with the only caller. buf_tmp_reserve_compression_buf(), buf_tmp_page_encrypt(), buf_page_encrypt(): Define static in the same compilation unit with the only caller. PageConverter::m_current_lsn: Remove. Write 0 to FIL_PAGE_LSN on ALTER TABLE...IMPORT TABLESPACE.
2019-11-25 08:39:51 +01:00
case TRX_SYS_PAGE_NO:
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
if (bpage->id().space() != TRX_SYS_SPACE)
MDEV-21132 Remove buf_page_t::newest_modification At each mini-transaction commit, the log sequence number of the mini-transaction must be written to each modified page, so that it will be available in the FIL_PAGE_LSN field when the page is being read in crash recovery. InnoDB was unnecessarily allocating redundant storage for the field, in buf_page_t::newest_modification. Let us access FIL_PAGE_LSN directly. Furthermore, on ALTER TABLE...IMPORT TABLESPACE, let us write 0 to FIL_PAGE_LSN instead of using log_sys.lsn. buf_flush_init_for_writing(), buf_flush_update_zip_checksum(), fil_encrypt_buf_for_full_crc32(), fil_encrypt_buf(), fil_space_encrypt(): Remove the parameter lsn. buf_page_get_newest_modification(): Merge with the only caller. buf_tmp_reserve_compression_buf(), buf_tmp_page_encrypt(), buf_page_encrypt(): Define static in the same compilation unit with the only caller. PageConverter::m_current_lsn: Remove. Write 0 to FIL_PAGE_LSN on ALTER TABLE...IMPORT TABLESPACE.
2019-11-25 08:39:51 +01:00
break;
/* The TRX_SYS page is neither encrypted nor compressed, because
it contains the address of the doublewrite buffer. */
/* fall through */
case 0:
/* Page 0 of a tablespace is not encrypted/compressed */
return s;
}
fil_space_crypt_t *crypt_data= space->crypt_data;
bool encrypted, page_compressed;
if (space->purpose == FIL_TYPE_TEMPORARY)
{
ut_ad(!crypt_data);
encrypted= innodb_encrypt_temporary_tables;
page_compressed= false;
}
else
{
encrypted= crypt_data && !crypt_data->not_encrypted() &&
crypt_data->type != CRYPT_SCHEME_UNENCRYPTED &&
(!crypt_data->is_default_encryption() || srv_encrypt_tables);
page_compressed= space->is_compressed();
}
const bool full_crc32= space->full_crc32();
if (!encrypted && !page_compressed)
{
/* No need to encrypt or compress. Clear key-version & crypt-checksum. */
MDEV-21133 Optimize access to InnoDB page header fields Introduce memcpy_aligned<N>(), memcmp_aligned<N>(), memset_aligned<N>() and use them for accessing InnoDB page header fields that are known to be aligned. MY_ASSUME_ALIGNED(): Wrapper for the GCC/clang __builtin_assume_aligned(). Nothing similar seems to exist in Microsoft Visual Studio, and the C++20 std::assume_aligned is not available to us yet. Explicitly specified alignment guarantees allow compilers to generate faster code on platforms with strict alignment rules, instead of emitting calls to potentially unaligned memcpy(), memcmp(), or memset().
2019-11-26 09:14:07 +01:00
static_assert(FIL_PAGE_FCRC32_KEY_VERSION % 4 == 0, "alignment");
static_assert(FIL_PAGE_FILE_FLUSH_LSN_OR_KEY_VERSION % 4 == 2,
"not perfect alignment");
MDEV-21132 Remove buf_page_t::newest_modification At each mini-transaction commit, the log sequence number of the mini-transaction must be written to each modified page, so that it will be available in the FIL_PAGE_LSN field when the page is being read in crash recovery. InnoDB was unnecessarily allocating redundant storage for the field, in buf_page_t::newest_modification. Let us access FIL_PAGE_LSN directly. Furthermore, on ALTER TABLE...IMPORT TABLESPACE, let us write 0 to FIL_PAGE_LSN instead of using log_sys.lsn. buf_flush_init_for_writing(), buf_flush_update_zip_checksum(), fil_encrypt_buf_for_full_crc32(), fil_encrypt_buf(), fil_space_encrypt(): Remove the parameter lsn. buf_page_get_newest_modification(): Merge with the only caller. buf_tmp_reserve_compression_buf(), buf_tmp_page_encrypt(), buf_page_encrypt(): Define static in the same compilation unit with the only caller. PageConverter::m_current_lsn: Remove. Write 0 to FIL_PAGE_LSN on ALTER TABLE...IMPORT TABLESPACE.
2019-11-25 08:39:51 +01:00
if (full_crc32)
MDEV-21133 Optimize access to InnoDB page header fields Introduce memcpy_aligned<N>(), memcmp_aligned<N>(), memset_aligned<N>() and use them for accessing InnoDB page header fields that are known to be aligned. MY_ASSUME_ALIGNED(): Wrapper for the GCC/clang __builtin_assume_aligned(). Nothing similar seems to exist in Microsoft Visual Studio, and the C++20 std::assume_aligned is not available to us yet. Explicitly specified alignment guarantees allow compilers to generate faster code on platforms with strict alignment rules, instead of emitting calls to potentially unaligned memcpy(), memcmp(), or memset().
2019-11-26 09:14:07 +01:00
memset_aligned<4>(s + FIL_PAGE_FCRC32_KEY_VERSION, 0, 4);
MDEV-21132 Remove buf_page_t::newest_modification At each mini-transaction commit, the log sequence number of the mini-transaction must be written to each modified page, so that it will be available in the FIL_PAGE_LSN field when the page is being read in crash recovery. InnoDB was unnecessarily allocating redundant storage for the field, in buf_page_t::newest_modification. Let us access FIL_PAGE_LSN directly. Furthermore, on ALTER TABLE...IMPORT TABLESPACE, let us write 0 to FIL_PAGE_LSN instead of using log_sys.lsn. buf_flush_init_for_writing(), buf_flush_update_zip_checksum(), fil_encrypt_buf_for_full_crc32(), fil_encrypt_buf(), fil_space_encrypt(): Remove the parameter lsn. buf_page_get_newest_modification(): Merge with the only caller. buf_tmp_reserve_compression_buf(), buf_tmp_page_encrypt(), buf_page_encrypt(): Define static in the same compilation unit with the only caller. PageConverter::m_current_lsn: Remove. Write 0 to FIL_PAGE_LSN on ALTER TABLE...IMPORT TABLESPACE.
2019-11-25 08:39:51 +01:00
else
MDEV-21133 Optimize access to InnoDB page header fields Introduce memcpy_aligned<N>(), memcmp_aligned<N>(), memset_aligned<N>() and use them for accessing InnoDB page header fields that are known to be aligned. MY_ASSUME_ALIGNED(): Wrapper for the GCC/clang __builtin_assume_aligned(). Nothing similar seems to exist in Microsoft Visual Studio, and the C++20 std::assume_aligned is not available to us yet. Explicitly specified alignment guarantees allow compilers to generate faster code on platforms with strict alignment rules, instead of emitting calls to potentially unaligned memcpy(), memcmp(), or memset().
2019-11-26 09:14:07 +01:00
memset_aligned<2>(s + FIL_PAGE_FILE_FLUSH_LSN_OR_KEY_VERSION, 0, 8);
MDEV-21132 Remove buf_page_t::newest_modification At each mini-transaction commit, the log sequence number of the mini-transaction must be written to each modified page, so that it will be available in the FIL_PAGE_LSN field when the page is being read in crash recovery. InnoDB was unnecessarily allocating redundant storage for the field, in buf_page_t::newest_modification. Let us access FIL_PAGE_LSN directly. Furthermore, on ALTER TABLE...IMPORT TABLESPACE, let us write 0 to FIL_PAGE_LSN instead of using log_sys.lsn. buf_flush_init_for_writing(), buf_flush_update_zip_checksum(), fil_encrypt_buf_for_full_crc32(), fil_encrypt_buf(), fil_space_encrypt(): Remove the parameter lsn. buf_page_get_newest_modification(): Merge with the only caller. buf_tmp_reserve_compression_buf(), buf_tmp_page_encrypt(), buf_page_encrypt(): Define static in the same compilation unit with the only caller. PageConverter::m_current_lsn: Remove. Write 0 to FIL_PAGE_LSN on ALTER TABLE...IMPORT TABLESPACE.
2019-11-25 08:39:51 +01:00
return s;
}
MDEV-21133 Optimize access to InnoDB page header fields Introduce memcpy_aligned<N>(), memcmp_aligned<N>(), memset_aligned<N>() and use them for accessing InnoDB page header fields that are known to be aligned. MY_ASSUME_ALIGNED(): Wrapper for the GCC/clang __builtin_assume_aligned(). Nothing similar seems to exist in Microsoft Visual Studio, and the C++20 std::assume_aligned is not available to us yet. Explicitly specified alignment guarantees allow compilers to generate faster code on platforms with strict alignment rules, instead of emitting calls to potentially unaligned memcpy(), memcmp(), or memset().
2019-11-26 09:14:07 +01:00
static_assert(FIL_PAGE_FCRC32_END_LSN % 4 == 0, "alignment");
static_assert(FIL_PAGE_LSN % 8 == 0, "alignment");
MDEV-21132 Remove buf_page_t::newest_modification At each mini-transaction commit, the log sequence number of the mini-transaction must be written to each modified page, so that it will be available in the FIL_PAGE_LSN field when the page is being read in crash recovery. InnoDB was unnecessarily allocating redundant storage for the field, in buf_page_t::newest_modification. Let us access FIL_PAGE_LSN directly. Furthermore, on ALTER TABLE...IMPORT TABLESPACE, let us write 0 to FIL_PAGE_LSN instead of using log_sys.lsn. buf_flush_init_for_writing(), buf_flush_update_zip_checksum(), fil_encrypt_buf_for_full_crc32(), fil_encrypt_buf(), fil_space_encrypt(): Remove the parameter lsn. buf_page_get_newest_modification(): Merge with the only caller. buf_tmp_reserve_compression_buf(), buf_tmp_page_encrypt(), buf_page_encrypt(): Define static in the same compilation unit with the only caller. PageConverter::m_current_lsn: Remove. Write 0 to FIL_PAGE_LSN on ALTER TABLE...IMPORT TABLESPACE.
2019-11-25 08:39:51 +01:00
if (full_crc32)
MDEV-21133 Optimize access to InnoDB page header fields Introduce memcpy_aligned<N>(), memcmp_aligned<N>(), memset_aligned<N>() and use them for accessing InnoDB page header fields that are known to be aligned. MY_ASSUME_ALIGNED(): Wrapper for the GCC/clang __builtin_assume_aligned(). Nothing similar seems to exist in Microsoft Visual Studio, and the C++20 std::assume_aligned is not available to us yet. Explicitly specified alignment guarantees allow compilers to generate faster code on platforms with strict alignment rules, instead of emitting calls to potentially unaligned memcpy(), memcmp(), or memset().
2019-11-26 09:14:07 +01:00
memcpy_aligned<4>(s + srv_page_size - FIL_PAGE_FCRC32_END_LSN,
FIL_PAGE_LSN + 4 + s, 4);
MDEV-21132 Remove buf_page_t::newest_modification At each mini-transaction commit, the log sequence number of the mini-transaction must be written to each modified page, so that it will be available in the FIL_PAGE_LSN field when the page is being read in crash recovery. InnoDB was unnecessarily allocating redundant storage for the field, in buf_page_t::newest_modification. Let us access FIL_PAGE_LSN directly. Furthermore, on ALTER TABLE...IMPORT TABLESPACE, let us write 0 to FIL_PAGE_LSN instead of using log_sys.lsn. buf_flush_init_for_writing(), buf_flush_update_zip_checksum(), fil_encrypt_buf_for_full_crc32(), fil_encrypt_buf(), fil_space_encrypt(): Remove the parameter lsn. buf_page_get_newest_modification(): Merge with the only caller. buf_tmp_reserve_compression_buf(), buf_tmp_page_encrypt(), buf_page_encrypt(): Define static in the same compilation unit with the only caller. PageConverter::m_current_lsn: Remove. Write 0 to FIL_PAGE_LSN on ALTER TABLE...IMPORT TABLESPACE.
2019-11-25 08:39:51 +01:00
ut_ad(!bpage->zip_size() || !page_compressed);
/* Find free slot from temporary memory array */
MDEV-27058: Move buf_page_t::slot to IORequest::slot MDEV-23855 and MDEV-23399 already moved some transient data fields from buffer pool page descriptors to IORequest, but the write buffer of PAGE_COMPRESSED or ENCRYPTED tables was missed. Since is only needed during asynchronous page write requests, it belongs to IORequest.
2021-11-16 17:16:53 +01:00
*slot= buf_pool.io_buf_reserve();
ut_a(*slot);
(*slot)->allocate();
MDEV-21132 Remove buf_page_t::newest_modification At each mini-transaction commit, the log sequence number of the mini-transaction must be written to each modified page, so that it will be available in the FIL_PAGE_LSN field when the page is being read in crash recovery. InnoDB was unnecessarily allocating redundant storage for the field, in buf_page_t::newest_modification. Let us access FIL_PAGE_LSN directly. Furthermore, on ALTER TABLE...IMPORT TABLESPACE, let us write 0 to FIL_PAGE_LSN instead of using log_sys.lsn. buf_flush_init_for_writing(), buf_flush_update_zip_checksum(), fil_encrypt_buf_for_full_crc32(), fil_encrypt_buf(), fil_space_encrypt(): Remove the parameter lsn. buf_page_get_newest_modification(): Merge with the only caller. buf_tmp_reserve_compression_buf(), buf_tmp_page_encrypt(), buf_page_encrypt(): Define static in the same compilation unit with the only caller. PageConverter::m_current_lsn: Remove. Write 0 to FIL_PAGE_LSN on ALTER TABLE...IMPORT TABLESPACE.
2019-11-25 08:39:51 +01:00
MDEV-27058: Move buf_page_t::slot to IORequest::slot MDEV-23855 and MDEV-23399 already moved some transient data fields from buffer pool page descriptors to IORequest, but the write buffer of PAGE_COMPRESSED or ENCRYPTED tables was missed. Since is only needed during asynchronous page write requests, it belongs to IORequest.
2021-11-16 17:16:53 +01:00
byte *d= (*slot)->crypt_buf;
MDEV-21132 Remove buf_page_t::newest_modification At each mini-transaction commit, the log sequence number of the mini-transaction must be written to each modified page, so that it will be available in the FIL_PAGE_LSN field when the page is being read in crash recovery. InnoDB was unnecessarily allocating redundant storage for the field, in buf_page_t::newest_modification. Let us access FIL_PAGE_LSN directly. Furthermore, on ALTER TABLE...IMPORT TABLESPACE, let us write 0 to FIL_PAGE_LSN instead of using log_sys.lsn. buf_flush_init_for_writing(), buf_flush_update_zip_checksum(), fil_encrypt_buf_for_full_crc32(), fil_encrypt_buf(), fil_space_encrypt(): Remove the parameter lsn. buf_page_get_newest_modification(): Merge with the only caller. buf_tmp_reserve_compression_buf(), buf_tmp_page_encrypt(), buf_page_encrypt(): Define static in the same compilation unit with the only caller. PageConverter::m_current_lsn: Remove. Write 0 to FIL_PAGE_LSN on ALTER TABLE...IMPORT TABLESPACE.
2019-11-25 08:39:51 +01:00
if (!page_compressed)
{
not_compressed:
MDEV-27058: Move buf_page_t::slot to IORequest::slot MDEV-23855 and MDEV-23399 already moved some transient data fields from buffer pool page descriptors to IORequest, but the write buffer of PAGE_COMPRESSED or ENCRYPTED tables was missed. Since is only needed during asynchronous page write requests, it belongs to IORequest.
2021-11-16 17:16:53 +01:00
d= space->purpose == FIL_TYPE_TEMPORARY
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
? buf_tmp_page_encrypt(page_no, s, d)
: fil_space_encrypt(space, page_no, s, d);
MDEV-21132 Remove buf_page_t::newest_modification At each mini-transaction commit, the log sequence number of the mini-transaction must be written to each modified page, so that it will be available in the FIL_PAGE_LSN field when the page is being read in crash recovery. InnoDB was unnecessarily allocating redundant storage for the field, in buf_page_t::newest_modification. Let us access FIL_PAGE_LSN directly. Furthermore, on ALTER TABLE...IMPORT TABLESPACE, let us write 0 to FIL_PAGE_LSN instead of using log_sys.lsn. buf_flush_init_for_writing(), buf_flush_update_zip_checksum(), fil_encrypt_buf_for_full_crc32(), fil_encrypt_buf(), fil_space_encrypt(): Remove the parameter lsn. buf_page_get_newest_modification(): Merge with the only caller. buf_tmp_reserve_compression_buf(), buf_tmp_page_encrypt(), buf_page_encrypt(): Define static in the same compilation unit with the only caller. PageConverter::m_current_lsn: Remove. Write 0 to FIL_PAGE_LSN on ALTER TABLE...IMPORT TABLESPACE.
2019-11-25 08:39:51 +01:00
}
else
{
ut_ad(space->purpose != FIL_TYPE_TEMPORARY);
/* First we compress the page content */
MDEV-27058: Move buf_page_t::slot to IORequest::slot MDEV-23855 and MDEV-23399 already moved some transient data fields from buffer pool page descriptors to IORequest, but the write buffer of PAGE_COMPRESSED or ENCRYPTED tables was missed. Since is only needed during asynchronous page write requests, it belongs to IORequest.
2021-11-16 17:16:53 +01:00
buf_tmp_reserve_compression_buf(*slot);
byte *tmp= (*slot)->comp_buf;
MDEV-21132 Remove buf_page_t::newest_modification At each mini-transaction commit, the log sequence number of the mini-transaction must be written to each modified page, so that it will be available in the FIL_PAGE_LSN field when the page is being read in crash recovery. InnoDB was unnecessarily allocating redundant storage for the field, in buf_page_t::newest_modification. Let us access FIL_PAGE_LSN directly. Furthermore, on ALTER TABLE...IMPORT TABLESPACE, let us write 0 to FIL_PAGE_LSN instead of using log_sys.lsn. buf_flush_init_for_writing(), buf_flush_update_zip_checksum(), fil_encrypt_buf_for_full_crc32(), fil_encrypt_buf(), fil_space_encrypt(): Remove the parameter lsn. buf_page_get_newest_modification(): Merge with the only caller. buf_tmp_reserve_compression_buf(), buf_tmp_page_encrypt(), buf_page_encrypt(): Define static in the same compilation unit with the only caller. PageConverter::m_current_lsn: Remove. Write 0 to FIL_PAGE_LSN on ALTER TABLE...IMPORT TABLESPACE.
2019-11-25 08:39:51 +01:00
ulint len= fil_page_compress(s, tmp, space->flags,
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
fil_space_get_block_size(space, page_no),
MDEV-21132 Remove buf_page_t::newest_modification At each mini-transaction commit, the log sequence number of the mini-transaction must be written to each modified page, so that it will be available in the FIL_PAGE_LSN field when the page is being read in crash recovery. InnoDB was unnecessarily allocating redundant storage for the field, in buf_page_t::newest_modification. Let us access FIL_PAGE_LSN directly. Furthermore, on ALTER TABLE...IMPORT TABLESPACE, let us write 0 to FIL_PAGE_LSN instead of using log_sys.lsn. buf_flush_init_for_writing(), buf_flush_update_zip_checksum(), fil_encrypt_buf_for_full_crc32(), fil_encrypt_buf(), fil_space_encrypt(): Remove the parameter lsn. buf_page_get_newest_modification(): Merge with the only caller. buf_tmp_reserve_compression_buf(), buf_tmp_page_encrypt(), buf_page_encrypt(): Define static in the same compilation unit with the only caller. PageConverter::m_current_lsn: Remove. Write 0 to FIL_PAGE_LSN on ALTER TABLE...IMPORT TABLESPACE.
2019-11-25 08:39:51 +01:00
encrypted);
if (!len)
goto not_compressed;
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
*size= len;
MDEV-21132 Remove buf_page_t::newest_modification At each mini-transaction commit, the log sequence number of the mini-transaction must be written to each modified page, so that it will be available in the FIL_PAGE_LSN field when the page is being read in crash recovery. InnoDB was unnecessarily allocating redundant storage for the field, in buf_page_t::newest_modification. Let us access FIL_PAGE_LSN directly. Furthermore, on ALTER TABLE...IMPORT TABLESPACE, let us write 0 to FIL_PAGE_LSN instead of using log_sys.lsn. buf_flush_init_for_writing(), buf_flush_update_zip_checksum(), fil_encrypt_buf_for_full_crc32(), fil_encrypt_buf(), fil_space_encrypt(): Remove the parameter lsn. buf_page_get_newest_modification(): Merge with the only caller. buf_tmp_reserve_compression_buf(), buf_tmp_page_encrypt(), buf_page_encrypt(): Define static in the same compilation unit with the only caller. PageConverter::m_current_lsn: Remove. Write 0 to FIL_PAGE_LSN on ALTER TABLE...IMPORT TABLESPACE.
2019-11-25 08:39:51 +01:00
if (full_crc32)
{
ut_d(bool compressed = false);
len= buf_page_full_crc32_size(tmp,
#ifdef UNIV_DEBUG
&compressed,
#else
NULL,
#endif
NULL);
ut_ad(compressed);
}
/* Workaround for MDEV-15527. */
memset(tmp + len, 0 , srv_page_size - len);
if (encrypted)
MDEV-27058: Move buf_page_t::slot to IORequest::slot MDEV-23855 and MDEV-23399 already moved some transient data fields from buffer pool page descriptors to IORequest, but the write buffer of PAGE_COMPRESSED or ENCRYPTED tables was missed. Since is only needed during asynchronous page write requests, it belongs to IORequest.
2021-11-16 17:16:53 +01:00
tmp= fil_space_encrypt(space, page_no, tmp, d);
MDEV-21132 Remove buf_page_t::newest_modification At each mini-transaction commit, the log sequence number of the mini-transaction must be written to each modified page, so that it will be available in the FIL_PAGE_LSN field when the page is being read in crash recovery. InnoDB was unnecessarily allocating redundant storage for the field, in buf_page_t::newest_modification. Let us access FIL_PAGE_LSN directly. Furthermore, on ALTER TABLE...IMPORT TABLESPACE, let us write 0 to FIL_PAGE_LSN instead of using log_sys.lsn. buf_flush_init_for_writing(), buf_flush_update_zip_checksum(), fil_encrypt_buf_for_full_crc32(), fil_encrypt_buf(), fil_space_encrypt(): Remove the parameter lsn. buf_page_get_newest_modification(): Merge with the only caller. buf_tmp_reserve_compression_buf(), buf_tmp_page_encrypt(), buf_page_encrypt(): Define static in the same compilation unit with the only caller. PageConverter::m_current_lsn: Remove. Write 0 to FIL_PAGE_LSN on ALTER TABLE...IMPORT TABLESPACE.
2019-11-25 08:39:51 +01:00
if (full_crc32)
{
MDEV-21133 Optimize access to InnoDB page header fields Introduce memcpy_aligned<N>(), memcmp_aligned<N>(), memset_aligned<N>() and use them for accessing InnoDB page header fields that are known to be aligned. MY_ASSUME_ALIGNED(): Wrapper for the GCC/clang __builtin_assume_aligned(). Nothing similar seems to exist in Microsoft Visual Studio, and the C++20 std::assume_aligned is not available to us yet. Explicitly specified alignment guarantees allow compilers to generate faster code on platforms with strict alignment rules, instead of emitting calls to potentially unaligned memcpy(), memcmp(), or memset().
2019-11-26 09:14:07 +01:00
static_assert(FIL_PAGE_FCRC32_CHECKSUM == 4, "alignment");
MDEV-21132 Remove buf_page_t::newest_modification At each mini-transaction commit, the log sequence number of the mini-transaction must be written to each modified page, so that it will be available in the FIL_PAGE_LSN field when the page is being read in crash recovery. InnoDB was unnecessarily allocating redundant storage for the field, in buf_page_t::newest_modification. Let us access FIL_PAGE_LSN directly. Furthermore, on ALTER TABLE...IMPORT TABLESPACE, let us write 0 to FIL_PAGE_LSN instead of using log_sys.lsn. buf_flush_init_for_writing(), buf_flush_update_zip_checksum(), fil_encrypt_buf_for_full_crc32(), fil_encrypt_buf(), fil_space_encrypt(): Remove the parameter lsn. buf_page_get_newest_modification(): Merge with the only caller. buf_tmp_reserve_compression_buf(), buf_tmp_page_encrypt(), buf_page_encrypt(): Define static in the same compilation unit with the only caller. PageConverter::m_current_lsn: Remove. Write 0 to FIL_PAGE_LSN on ALTER TABLE...IMPORT TABLESPACE.
2019-11-25 08:39:51 +01:00
mach_write_to_4(tmp + len - 4, ut_crc32(tmp, len - 4));
ut_ad(!buf_page_is_corrupted(true, tmp, space->flags));
}
MDEV-27058: Move buf_page_t::slot to IORequest::slot MDEV-23855 and MDEV-23399 already moved some transient data fields from buffer pool page descriptors to IORequest, but the write buffer of PAGE_COMPRESSED or ENCRYPTED tables was missed. Since is only needed during asynchronous page write requests, it belongs to IORequest.
2021-11-16 17:16:53 +01:00
d= tmp;
MDEV-21132 Remove buf_page_t::newest_modification At each mini-transaction commit, the log sequence number of the mini-transaction must be written to each modified page, so that it will be available in the FIL_PAGE_LSN field when the page is being read in crash recovery. InnoDB was unnecessarily allocating redundant storage for the field, in buf_page_t::newest_modification. Let us access FIL_PAGE_LSN directly. Furthermore, on ALTER TABLE...IMPORT TABLESPACE, let us write 0 to FIL_PAGE_LSN instead of using log_sys.lsn. buf_flush_init_for_writing(), buf_flush_update_zip_checksum(), fil_encrypt_buf_for_full_crc32(), fil_encrypt_buf(), fil_space_encrypt(): Remove the parameter lsn. buf_page_get_newest_modification(): Merge with the only caller. buf_tmp_reserve_compression_buf(), buf_tmp_page_encrypt(), buf_page_encrypt(): Define static in the same compilation unit with the only caller. PageConverter::m_current_lsn: Remove. Write 0 to FIL_PAGE_LSN on ALTER TABLE...IMPORT TABLESPACE.
2019-11-25 08:39:51 +01:00
}
MDEV-27058: Move buf_page_t::slot to IORequest::slot MDEV-23855 and MDEV-23399 already moved some transient data fields from buffer pool page descriptors to IORequest, but the write buffer of PAGE_COMPRESSED or ENCRYPTED tables was missed. Since is only needed during asynchronous page write requests, it belongs to IORequest.
2021-11-16 17:16:53 +01:00
(*slot)->out_buf= d;
MDEV-21132 Remove buf_page_t::newest_modification At each mini-transaction commit, the log sequence number of the mini-transaction must be written to each modified page, so that it will be available in the FIL_PAGE_LSN field when the page is being read in crash recovery. InnoDB was unnecessarily allocating redundant storage for the field, in buf_page_t::newest_modification. Let us access FIL_PAGE_LSN directly. Furthermore, on ALTER TABLE...IMPORT TABLESPACE, let us write 0 to FIL_PAGE_LSN instead of using log_sys.lsn. buf_flush_init_for_writing(), buf_flush_update_zip_checksum(), fil_encrypt_buf_for_full_crc32(), fil_encrypt_buf(), fil_space_encrypt(): Remove the parameter lsn. buf_page_get_newest_modification(): Merge with the only caller. buf_tmp_reserve_compression_buf(), buf_tmp_page_encrypt(), buf_page_encrypt(): Define static in the same compilation unit with the only caller. PageConverter::m_current_lsn: Remove. Write 0 to FIL_PAGE_LSN on ALTER TABLE...IMPORT TABLESPACE.
2019-11-25 08:39:51 +01:00
return d;
}
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
/** Free a page whose underlying file page has been freed. */
inline void buf_pool_t::release_freed_page(buf_page_t *bpage)
MDEV-15528 Punch holes when pages are freed When a InnoDB data file page is freed, its contents becomes garbage, and any storage allocated in the data file is wasted. During flushing, InnoDB initializes the page with zeros if scrubbing is enabled. If the tablespace is compressed then InnoDB should punch a hole else ignore the flushing of the freed page. buf_page_t: - Replaced the variable file_page_was_freed, init_on_flush in buf_page_t with status enum variable. - Changed all debug assert of file_page_was_freed to DBUG_ASSERT of buf_page_t::status Removed buf_page_set_file_page_was_freed(), buf_page_reset_file_page_was_freed(). buf_page_free(): Newly added function which takes X-lock on the page before marking the status as FREED. So that InnoDB flush handler can avoid concurrent flush of the freed page. Also while flushing the page, InnoDB make sure that redo log which does freeing of the page also written to the disk. Currently, this function only marks the page as FREED if it is in buffer pool buf_flush_freed_page(): Newly added function which initializes zeros asynchorously if innodb_immediate_scrub_data_uncompressed is enabled. Punch a hole to the file synchorously if page_compressed is enabled. Reset the io_fix to NORMAL. Release the block from flush list and associated mutex before writing zeros or punch a hole to the file. buf_flush_page(): Removed the unnecessary usage of temporary variable "flush" fil_io(): Introduce new parameter called punch_hole. It allows fil_io() to punch the hole to the file for the given offset. buf_page_create(): Let the callers assign buf_page_t::status. Every caller should eventually invoke mtr_t::init(). fsp_page_create(): Remove the unused mtr_t parameter. In all other callers of buf_page_create() except fsp_page_create(), before invoking mtr_t::init(), invoke mtr_t::sx_latch_at_savepoint() or mtr_t::x_latch_at_savepoint(). mtr_t::init(): Initialize buf_page_t::status also for the temporary tablespace (when redo logging is disabled), to avoid assertion failures.
2020-03-09 08:55:33 +01:00
{
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
mysql_mutex_assert_owner(&mutex);
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
mysql_mutex_lock(&flush_list_mutex);
MDEV-26010 fixup: Use acquire/release memory order In commit 5f22511e35674ecc376f4c56217ee2d78f92c772 we depend on Total Store Ordering. For correct operation on ISAs that implement weaker memory ordering, we must explicitly use release/acquire stores and loads on buf_page_t::oldest_modification_ to prevent a race condition when buf_page_t::list does not happen to be on the same cache line. buf_page_t::clear_oldest_modification(): Assert that the block is not in buf_pool.flush_list, and use std::memory_order_release. buf_page_t::oldest_modification_acquire(): Read oldest_modification_ with std::memory_order_acquire. In this way, if the return value is 0, the caller may safely assume that it will not observe the buf_page_t as being in buf_pool.flush_list, even if it is not holding buf_pool.flush_list_mutex. buf_flush_relocate_on_flush_list(), buf_LRU_free_page(): Invoke buf_page_t::oldest_modification_acquire().
2021-06-26 10:16:40 +02:00
ut_d(const lsn_t oldest_modification= bpage->oldest_modification();)
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
if (fsp_is_system_temporary(bpage->id().space()))
{
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
ut_ad(bpage->frame);
MDEV-26010 fixup: Use acquire/release memory order In commit 5f22511e35674ecc376f4c56217ee2d78f92c772 we depend on Total Store Ordering. For correct operation on ISAs that implement weaker memory ordering, we must explicitly use release/acquire stores and loads on buf_page_t::oldest_modification_ to prevent a race condition when buf_page_t::list does not happen to be on the same cache line. buf_page_t::clear_oldest_modification(): Assert that the block is not in buf_pool.flush_list, and use std::memory_order_release. buf_page_t::oldest_modification_acquire(): Read oldest_modification_ with std::memory_order_acquire. In this way, if the return value is 0, the caller may safely assume that it will not observe the buf_page_t as being in buf_pool.flush_list, even if it is not holding buf_pool.flush_list_mutex. buf_flush_relocate_on_flush_list(), buf_LRU_free_page(): Invoke buf_page_t::oldest_modification_acquire().
2021-06-26 10:16:40 +02:00
ut_ad(oldest_modification == 2);
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
}
else
{
MDEV-26010 fixup: Use acquire/release memory order In commit 5f22511e35674ecc376f4c56217ee2d78f92c772 we depend on Total Store Ordering. For correct operation on ISAs that implement weaker memory ordering, we must explicitly use release/acquire stores and loads on buf_page_t::oldest_modification_ to prevent a race condition when buf_page_t::list does not happen to be on the same cache line. buf_page_t::clear_oldest_modification(): Assert that the block is not in buf_pool.flush_list, and use std::memory_order_release. buf_page_t::oldest_modification_acquire(): Read oldest_modification_ with std::memory_order_acquire. In this way, if the return value is 0, the caller may safely assume that it will not observe the buf_page_t as being in buf_pool.flush_list, even if it is not holding buf_pool.flush_list_mutex. buf_flush_relocate_on_flush_list(), buf_LRU_free_page(): Invoke buf_page_t::oldest_modification_acquire().
2021-06-26 10:16:40 +02:00
ut_ad(oldest_modification > 2);
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
delete_from_flush_list(bpage, false);
}
MDEV-12227 Defer writes to the InnoDB temporary tablespace The flushing of the InnoDB temporary tablespace is unnecessarily tied to the write-ahead redo logging and redo log checkpoints, which must be tied to the page writes of persistent tablespaces. Let us simply omit any pages of temporary tables from buf_pool.flush_list. In this way, log checkpoints will never incur any 'collateral damage' of writing out unmodified changes for temporary tables. After this change, pages of the temporary tablespace can only be written out by buf_flush_lists(n_pages,0) as part of LRU eviction. Hopefully, most of the time, that code will never be executed, and instead, the temporary pages will be evicted by buf_release_freed_page() without ever being written back to the temporary tablespace file. This should improve the efficiency of the checkpoint flushing and the buf_flush_page_cleaner thread. Reviewed by: Vladislav Vaintroub
2020-12-09 08:22:13 +01:00
bpage->clear_oldest_modification();
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
mysql_mutex_unlock(&flush_list_mutex);
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
bpage->lock.u_unlock(true);
MDEV-15528: Minor cleanup buf_flush_freed_page(): Reformat in the common style, and simplify some code. Prefer to request all information from smaller data structures (buf_page_t) than from fil_space_t or the global variable srv_immediate_scrub_data_uncompressed. SysTablespace::open_or_create(): Assert that the temporary tablespace will not be created in page_compressed format, so that buf_flush_freed_page() can avoid checking that on every call. IORequest: Remove duplicated constructors, and do not explicitly declare a default constructor.
2020-03-10 08:34:09 +01:00
MDEV-8139 Fix Scrubbing fil_space_t::freed_ranges: Store ranges of freed page numbers. fil_space_t::last_freed_lsn: Store the most recent LSN of freeing a page. fil_space_t::freed_mutex: Protects freed_ranges, last_freed_lsn. fil_space_create(): Initialize the freed_range mutex. fil_space_free_low(): Frees the freed_range mutex. range_set: Ranges of page numbers. buf_page_create(): Removes the page from freed_ranges when page is being reused. btr_free_root(): Remove the PAGE_INDEX_ID invalidation. Because btr_free_root() and dict_drop_index_tree() are executed in the same atomic mini-transaction, there is no need to invalidate the root page. buf_release_freed_page(): Split from buf_flush_freed_page(). Skip any I/O buf_flush_freed_pages(): Get the freed ranges from tablespace and Write punch-hole or zeroes of the freed ranges. buf_flush_try_neighbors(): Handles the flushing of freed ranges. mtr_t::freed_pages: Variable to store the list of freed pages. mtr_t::add_freed_pages(): To add freed pages. mtr_t::clear_freed_pages(): To clear the freed pages. mtr_t::m_freed_in_system_tablespace: Variable to indicate whether page has been freed in system tablespace. mtr_t::m_trim_pages: Variable to indicate whether the space has been trimmed. mtr_t::commit(): Add the freed page and update the last freed lsn in the tablespace and clear the tablespace freed range if space is trimmed. file_name_t::freed_pages: Store the freed pages during recovery. file_name_t::add_freed_page(), file_name_t::remove_freed_page(): To add and remove freed page during recovery. store_freed_or_init_rec(): Store or remove the freed pages while encountering FREE_PAGE or INIT_PAGE redo log record. recv_init_crash_recovery_spaces(): Add the freed page encountered during recovery to respective tablespace.
2020-06-11 19:22:47 +02:00
buf_LRU_free_page(bpage, true);
MDEV-15528 Punch holes when pages are freed When a InnoDB data file page is freed, its contents becomes garbage, and any storage allocated in the data file is wasted. During flushing, InnoDB initializes the page with zeros if scrubbing is enabled. If the tablespace is compressed then InnoDB should punch a hole else ignore the flushing of the freed page. buf_page_t: - Replaced the variable file_page_was_freed, init_on_flush in buf_page_t with status enum variable. - Changed all debug assert of file_page_was_freed to DBUG_ASSERT of buf_page_t::status Removed buf_page_set_file_page_was_freed(), buf_page_reset_file_page_was_freed(). buf_page_free(): Newly added function which takes X-lock on the page before marking the status as FREED. So that InnoDB flush handler can avoid concurrent flush of the freed page. Also while flushing the page, InnoDB make sure that redo log which does freeing of the page also written to the disk. Currently, this function only marks the page as FREED if it is in buffer pool buf_flush_freed_page(): Newly added function which initializes zeros asynchorously if innodb_immediate_scrub_data_uncompressed is enabled. Punch a hole to the file synchorously if page_compressed is enabled. Reset the io_fix to NORMAL. Release the block from flush list and associated mutex before writing zeros or punch a hole to the file. buf_flush_page(): Removed the unnecessary usage of temporary variable "flush" fil_io(): Introduce new parameter called punch_hole. It allows fil_io() to punch the hole to the file for the given offset. buf_page_create(): Let the callers assign buf_page_t::status. Every caller should eventually invoke mtr_t::init(). fsp_page_create(): Remove the unused mtr_t parameter. In all other callers of buf_page_create() except fsp_page_create(), before invoking mtr_t::init(), invoke mtr_t::sx_latch_at_savepoint() or mtr_t::x_latch_at_savepoint(). mtr_t::init(): Initialize buf_page_t::status also for the temporary tablespace (when redo logging is disabled), to avoid assertion failures.
2020-03-09 08:55:33 +01:00
}
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
/** Write a flushable page to a file. buf_pool.mutex must be held.
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
@param lru true=buf_pool.LRU; false=buf_pool.flush_list
@param space tablespace
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
@return whether the page was flushed and buf_pool.mutex was released */
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
inline bool buf_page_t::flush(bool lru, fil_space_t *space)
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
{
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
ut_ad(in_file());
ut_ad(in_LRU_list);
MDEV-23855: Remove fil_system.LRU and reduce fil_system.mutex contention Also fixes MDEV-23929: innodb_flush_neighbors is not being ignored for system tablespace on SSD When the maximum configured number of file is exceeded, InnoDB will close data files. We used to maintain a fil_system.LRU list and a counter fil_node_t::n_pending to achieve this, at the huge cost of multiple fil_system.mutex operations per I/O operation. fil_node_open_file_low(): Implement a FIFO replacement policy: The last opened file will be moved to the end of fil_system.space_list, and files will be closed from the start of the list. However, we will not move tablespaces in fil_system.space_list while i_s_tablespaces_encryption_fill_table() is executing (producing output for INFORMATION_SCHEMA.INNODB_TABLESPACES_ENCRYPTION) because it may cause information of some tablespaces to go missing. We also avoid this in mariabackup --backup because datafiles_iter_next() assumes that the ordering is not changed. IORequest: Fold more parameters to IORequest::type. fil_space_t::io(): Replaces fil_io(). fil_space_t::flush(): Replaces fil_flush(). OS_AIO_IBUF: Remove. We will always issue synchronous reads of the change buffer pages in buf_read_page_low(). We will always ignore some errors for background reads. This should reduce fil_system.mutex contention a little. fil_node_t::complete_write(): Replaces fil_node_t::complete_io(). On both read and write completion, fil_space_t::release_for_io() will have to be called. fil_space_t::io(): Do not acquire fil_system.mutex in the normal code path. xb_delta_open_matching_space(): Do not try to open the system tablespace which was already opened. This fixes a file sharing violation in mariabackup --prepare --incremental. Reviewed by: Vladislav Vaintroub
2020-10-26 14:59:30 +01:00
ut_ad((space->purpose == FIL_TYPE_TEMPORARY) ==
(space == fil_system.temp_space));
MDEV-23855: Shrink fil_space_t Merge n_pending_ios, n_pending_ops to std::atomic<uint32_t> n_pending. Change some more fil_space_t members to uint32_t to reduce the memory footprint. fil_space_t::add(), fil_ibd_create(): Attach the already opened handle to the tablespace, and enforce the fil_system.n_open limit. dict_boot(): Initialize fil_system.max_assigned_id. srv_boot(): Call srv_thread_pool_init() before anything else, so that files should be opened in the correct mode on Windows. fil_ibd_create(): Create the file in OS_FILE_AIO mode, just like fil_node_open_file_low() does it. dict_table_t::is_accessible(): Replaces fil_table_accessible(). Reviewed by: Vladislav Vaintroub
2020-10-26 15:04:12 +01:00
ut_ad(space->referenced());
MDEV-26010 fixup: Use acquire/release memory order In commit 5f22511e35674ecc376f4c56217ee2d78f92c772 we depend on Total Store Ordering. For correct operation on ISAs that implement weaker memory ordering, we must explicitly use release/acquire stores and loads on buf_page_t::oldest_modification_ to prevent a race condition when buf_page_t::list does not happen to be on the same cache line. buf_page_t::clear_oldest_modification(): Assert that the block is not in buf_pool.flush_list, and use std::memory_order_release. buf_page_t::oldest_modification_acquire(): Read oldest_modification_ with std::memory_order_acquire. In this way, if the return value is 0, the caller may safely assume that it will not observe the buf_page_t as being in buf_pool.flush_list, even if it is not holding buf_pool.flush_list_mutex. buf_flush_relocate_on_flush_list(), buf_LRU_free_page(): Invoke buf_page_t::oldest_modification_acquire().
2021-06-26 10:16:40 +02:00
ut_ad(lru || space != fil_system.temp_space);
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
if (!lock.u_lock_try(true))
return false;
const auto s= state();
ut_a(s >= FREED);
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
if (s < UNFIXED)
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
{
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
buf_pool.release_freed_page(this);
mysql_mutex_unlock(&buf_pool.mutex);
return true;
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
}
MDEV-21132 Remove buf_page_t::newest_modification At each mini-transaction commit, the log sequence number of the mini-transaction must be written to each modified page, so that it will be available in the FIL_PAGE_LSN field when the page is being read in crash recovery. InnoDB was unnecessarily allocating redundant storage for the field, in buf_page_t::newest_modification. Let us access FIL_PAGE_LSN directly. Furthermore, on ALTER TABLE...IMPORT TABLESPACE, let us write 0 to FIL_PAGE_LSN instead of using log_sys.lsn. buf_flush_init_for_writing(), buf_flush_update_zip_checksum(), fil_encrypt_buf_for_full_crc32(), fil_encrypt_buf(), fil_space_encrypt(): Remove the parameter lsn. buf_page_get_newest_modification(): Merge with the only caller. buf_tmp_reserve_compression_buf(), buf_tmp_page_encrypt(), buf_page_encrypt(): Define static in the same compilation unit with the only caller. PageConverter::m_current_lsn: Remove. Write 0 to FIL_PAGE_LSN on ALTER TABLE...IMPORT TABLESPACE.
2019-11-25 08:39:51 +01:00
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
if (s >= READ_FIX || oldest_modification() < 2)
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
{
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
lock.u_unlock(true);
return false;
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
}
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
mysql_mutex_assert_not_owner(&buf_pool.flush_list_mutex);
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
/* Apart from the U-lock, this block will also be protected by
is_write_fixed() and oldest_modification()>1.
Thus, it cannot be relocated or removed. */
MDEV-12545 Reduce the amount of fil_space_t lookups buf_flush_write_block_low(): Acquire the tablespace reference once, and pass it to lower-level functions. This is only a start; further calls may be removed. fil_decompress_page(): Remove unsafe use of fil_space_get_by_id().
2017-04-21 10:28:18 +02:00
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
DBUG_PRINT("ib_buf", ("%s %u page %u:%u",
lru ? "LRU" : "flush_list",
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
id().space(), id().page_no()));
ut_d(const auto f=) zip.fix.fetch_add(WRITE_FIX - UNFIXED);
ut_ad(f >= UNFIXED);
ut_ad(f < READ_FIX);
MDEV-26010 fixup: Use acquire/release memory order In commit 5f22511e35674ecc376f4c56217ee2d78f92c772 we depend on Total Store Ordering. For correct operation on ISAs that implement weaker memory ordering, we must explicitly use release/acquire stores and loads on buf_page_t::oldest_modification_ to prevent a race condition when buf_page_t::list does not happen to be on the same cache line. buf_page_t::clear_oldest_modification(): Assert that the block is not in buf_pool.flush_list, and use std::memory_order_release. buf_page_t::oldest_modification_acquire(): Read oldest_modification_ with std::memory_order_acquire. In this way, if the return value is 0, the caller may safely assume that it will not observe the buf_page_t as being in buf_pool.flush_list, even if it is not holding buf_pool.flush_list_mutex. buf_flush_relocate_on_flush_list(), buf_LRU_free_page(): Invoke buf_page_t::oldest_modification_acquire().
2021-06-26 10:16:40 +02:00
ut_ad(space == fil_system.temp_space
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
? oldest_modification() == 2
: oldest_modification() > 2);
if (lru)
{
ut_ad(buf_pool.n_flush_LRU_ < ULINT_UNDEFINED);
buf_pool.n_flush_LRU_++;
}
else
{
ut_ad(buf_pool.n_flush_list_ < ULINT_UNDEFINED);
buf_pool.n_flush_list_++;
}
buf_flush_page_count++;
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
mysql_mutex_unlock(&buf_pool.mutex);
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
buf_block_t *block= reinterpret_cast<buf_block_t*>(this);
page_t *write_frame= zip.data;
space->reacquire();
size_t size;
#if defined HAVE_FALLOC_PUNCH_HOLE_AND_KEEP_SIZE || defined _WIN32
size_t orig_size;
#endif
IORequest::Type type= lru ? IORequest::WRITE_LRU : IORequest::WRITE_ASYNC;
buf_tmp_buffer_t *slot= nullptr;
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
if (UNIV_UNLIKELY(!frame)) /* ROW_FORMAT=COMPRESSED */
{
ut_ad(!space->full_crc32());
ut_ad(!space->is_compressed()); /* not page_compressed */
size= zip_size();
#if defined HAVE_FALLOC_PUNCH_HOLE_AND_KEEP_SIZE || defined _WIN32
orig_size= size;
#endif
buf_flush_update_zip_checksum(write_frame, size);
write_frame= buf_page_encrypt(space, this, write_frame, &slot, &size);
ut_ad(size == zip_size());
}
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
else
{
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
byte *page= frame;
size= block->physical_size();
Fix -Wunused-but-set-variable
2020-12-08 18:05:08 +01:00
#if defined HAVE_FALLOC_PUNCH_HOLE_AND_KEEP_SIZE || defined _WIN32
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
orig_size= size;
Fix -Wunused-but-set-variable
2020-12-08 18:05:08 +01:00
#endif
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
if (space->full_crc32())
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
{
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
/* innodb_checksum_algorithm=full_crc32 is not implemented for
ROW_FORMAT=COMPRESSED pages. */
ut_ad(!write_frame);
page= buf_page_encrypt(space, this, page, &slot, &size);
buf_flush_init_for_writing(block, page, nullptr, true);
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
}
else
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
{
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
buf_flush_init_for_writing(block, page, write_frame ? &zip : nullptr,
false);
page= buf_page_encrypt(space, this, write_frame ? write_frame : page,
&slot, &size);
}
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
#if defined HAVE_FALLOC_PUNCH_HOLE_AND_KEEP_SIZE || defined _WIN32
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
if (size != orig_size)
{
switch (space->chain.start->punch_hole) {
case 1:
type= lru ? IORequest::PUNCH_LRU : IORequest::PUNCH;
break;
case 2:
size= orig_size;
MDEV-26029: Sparse files are inefficient on thinly provisioned storage The MariaDB implementation of page_compressed tables for InnoDB used sparse files. In the worst case, in the data file, every data page will consist of some data followed by a hole. This may be extremely inefficient in some file systems. If the underlying storage device is thinly provisioned (can compress data on the fly), it would be good to write regular files (with sequences of NUL bytes at the end of each page_compressed block) and let the storage device take care of compressing the data. For reads, sparse file regions and regions containing NUL bytes will be indistinguishable. my_test_if_disable_punch_hole(): A new predicate for detecting thinly provisioned storage. (Not implemented yet.) innodb_atomic_writes: Correct the comment. buf_flush_page(): Support all values of fil_node_t::punch_hole. On a thinly provisioned storage device, we will always write NUL-padded innodb_page_size bytes also for page_compressed tables. buf_flush_freed_pages(): Remove a redundant condition. fil_space_t::atomic_write_supported: Remove. (This was duplicating fil_node_t::atomic_write.) fil_space_t::punch_hole: Remove. (Duplicated fil_node_t::punch_hole.) fil_node_t: Remove magic_n, and consolidate flags into bitfields. For punch_hole we introduce a third value that indicates a thinly provisioned storage device. fil_node_t::find_metadata(): Detect all attributes of the file.
2021-06-28 12:42:43 +02:00
}
MDEV-23855: Remove fil_system.LRU and reduce fil_system.mutex contention Also fixes MDEV-23929: innodb_flush_neighbors is not being ignored for system tablespace on SSD When the maximum configured number of file is exceeded, InnoDB will close data files. We used to maintain a fil_system.LRU list and a counter fil_node_t::n_pending to achieve this, at the huge cost of multiple fil_system.mutex operations per I/O operation. fil_node_open_file_low(): Implement a FIFO replacement policy: The last opened file will be moved to the end of fil_system.space_list, and files will be closed from the start of the list. However, we will not move tablespaces in fil_system.space_list while i_s_tablespaces_encryption_fill_table() is executing (producing output for INFORMATION_SCHEMA.INNODB_TABLESPACES_ENCRYPTION) because it may cause information of some tablespaces to go missing. We also avoid this in mariabackup --backup because datafiles_iter_next() assumes that the ordering is not changed. IORequest: Fold more parameters to IORequest::type. fil_space_t::io(): Replaces fil_io(). fil_space_t::flush(): Replaces fil_flush(). OS_AIO_IBUF: Remove. We will always issue synchronous reads of the change buffer pages in buf_read_page_low(). We will always ignore some errors for background reads. This should reduce fil_system.mutex contention a little. fil_node_t::complete_write(): Replaces fil_node_t::complete_io(). On both read and write completion, fil_space_t::release_for_io() will have to be called. fil_space_t::io(): Do not acquire fil_system.mutex in the normal code path. xb_delta_open_matching_space(): Do not try to open the system tablespace which was already opened. This fixes a file sharing violation in mariabackup --prepare --incremental. Reviewed by: Vladislav Vaintroub
2020-10-26 14:59:30 +01:00
}
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
#endif
write_frame= page;
}
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
if ((s & LRU_MASK) == REINIT || !space->use_doublewrite())
{
if (UNIV_LIKELY(space->purpose == FIL_TYPE_TABLESPACE))
MDEV-25948 Remove log_flush_task Vladislav Vaintroub suggested that invoking log_flush_up_to() for every page could perform better than invoking a log write between buf_pool.flush_list batches, like we started doing in commit 3a9a3be1c64b14c05648e87ebe0f1dd96457de41 (MDEV-23855). This could depend on the sequence in which pages are being modified. The buf_pool.flush_list is ordered by oldest_modification, while the FIL_PAGE_LSN of the pages is theoretically independent of that. In the pathological case, we will wait for a log write before writing each individual page. It turns out that we can defer the call to log_flush_up_to() until just before submitting the page write. If the doublewrite buffer is being used, we can submit a write batch of "future" pages to the doublewrite buffer, and only wait for the log write right before we are writing an already doublewritten page. The next doublewrite batch will not be initiated before the last page write from the current batch has completed. When a future version introduces asynchronous writes if the log, we could initiate a write at the start of a flushing batch, to reduce waiting further.
2021-06-23 11:01:41 +02:00
{
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
const lsn_t lsn=
mach_read_from_8(my_assume_aligned<8>(FIL_PAGE_LSN +
(write_frame ? write_frame
: frame)));
ut_ad(lsn >= oldest_modification());
if (lsn > log_sys.get_flushed_lsn())
log_write_up_to(lsn, true);
MDEV-25948 Remove log_flush_task Vladislav Vaintroub suggested that invoking log_flush_up_to() for every page could perform better than invoking a log write between buf_pool.flush_list batches, like we started doing in commit 3a9a3be1c64b14c05648e87ebe0f1dd96457de41 (MDEV-23855). This could depend on the sequence in which pages are being modified. The buf_pool.flush_list is ordered by oldest_modification, while the FIL_PAGE_LSN of the pages is theoretically independent of that. In the pathological case, we will wait for a log write before writing each individual page. It turns out that we can defer the call to log_flush_up_to() until just before submitting the page write. If the doublewrite buffer is being used, we can submit a write batch of "future" pages to the doublewrite buffer, and only wait for the log write right before we are writing an already doublewritten page. The next doublewrite batch will not be initiated before the last page write from the current batch has completed. When a future version introduces asynchronous writes if the log, we could initiate a write at the start of a flushing batch, to reduce waiting further.
2021-06-23 11:01:41 +02:00
}
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
space->io(IORequest{type, this, slot}, physical_offset(), size,
write_frame, this);
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
}
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
else
buf_dblwr.add_to_batch(IORequest{this, slot, space->chain.start, type},
size);
MDEV-15528 Punch holes when pages are freed When a InnoDB data file page is freed, its contents becomes garbage, and any storage allocated in the data file is wasted. During flushing, InnoDB initializes the page with zeros if scrubbing is enabled. If the tablespace is compressed then InnoDB should punch a hole else ignore the flushing of the freed page. buf_page_t: - Replaced the variable file_page_was_freed, init_on_flush in buf_page_t with status enum variable. - Changed all debug assert of file_page_was_freed to DBUG_ASSERT of buf_page_t::status Removed buf_page_set_file_page_was_freed(), buf_page_reset_file_page_was_freed(). buf_page_free(): Newly added function which takes X-lock on the page before marking the status as FREED. So that InnoDB flush handler can avoid concurrent flush of the freed page. Also while flushing the page, InnoDB make sure that redo log which does freeing of the page also written to the disk. Currently, this function only marks the page as FREED if it is in buffer pool buf_flush_freed_page(): Newly added function which initializes zeros asynchorously if innodb_immediate_scrub_data_uncompressed is enabled. Punch a hole to the file synchorously if page_compressed is enabled. Reset the io_fix to NORMAL. Release the block from flush list and associated mutex before writing zeros or punch a hole to the file. buf_flush_page(): Removed the unnecessary usage of temporary variable "flush" fil_io(): Introduce new parameter called punch_hole. It allows fil_io() to punch the hole to the file for the given offset. buf_page_create(): Let the callers assign buf_page_t::status. Every caller should eventually invoke mtr_t::init(). fsp_page_create(): Remove the unused mtr_t parameter. In all other callers of buf_page_create() except fsp_page_create(), before invoking mtr_t::init(), invoke mtr_t::sx_latch_at_savepoint() or mtr_t::x_latch_at_savepoint(). mtr_t::init(): Initialize buf_page_t::status also for the temporary tablespace (when redo logging is disabled), to avoid assertion failures.
2020-03-09 08:55:33 +01:00
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
/* Increment the I/O operation count used for selecting LRU policy. */
buf_LRU_stat_inc_io();
return true;
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
}
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
/** Check whether a page can be flushed from the buf_pool.
@param id page identifier
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
@param fold id.fold()
@param lru true=buf_pool.LRU; false=buf_pool.flush_list
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
@return whether the page can be flushed */
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
static bool buf_flush_check_neighbor(const page_id_t id, ulint fold, bool lru)
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
{
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
mysql_mutex_assert_owner(&buf_pool.mutex);
ut_ad(fold == id.fold());
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-26826 Duplicated computations of buf_pool.page_hash addresses Since commit bd5a6403cace36c6ed428cde62e35adcd3f7e7d0 (MDEV-26033) we can actually calculate the buf_pool.page_hash cell and latch addresses while not holding buf_pool.mutex. buf_page_alloc_descriptor(): Remove the MEM_UNDEFINED. We now expect buf_page_t::hash to be zero-initialized. buf_pool_t::hash_chain: Dedicated data type for buf_pool.page_hash.array. buf_LRU_free_one_page(): Merged to the only caller buf_pool_t::corrupted_evict().
2021-10-22 11:33:37 +02:00
/* FIXME: cell_get() is being invoked while holding buf_pool.mutex */
const buf_page_t *bpage=
buf_pool.page_hash.get(id, buf_pool.page_hash.cell_get(fold));
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
if (!bpage || buf_pool.watch_is_sentinel(*bpage))
return false;
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
/* We avoid flushing 'non-old' blocks in an LRU flush, because the
flushed blocks are soon freed */
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
if (lru && !bpage->is_old())
return false;
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
return bpage->oldest_modification() > 1 && bpage->ready_for_flush();
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
}
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
/** Check which neighbors of a page can be flushed from the buf_pool.
@param space tablespace
@param id page identifier of a dirty page
MDEV-23909 innodb_flush_neighbors=2 is treated like innodb_flush_neighbors=0 In MDEV-15053 (commit b1ab211dee599eabd9a5b886fafa3adea29ae041) we inadvertently removed a check whether innodb_flush_neighbors is 0, and thus started treating only the value 1 in a special way. buf_flush_check_neighbors(): Add the parameter contiguous, which can be set to skip the check for non-contiguous page number ranges. Reviewed by: Thirunarayanan Balathandayuthapani
2020-10-08 10:13:47 +02:00
@param contiguous whether to consider contiguous areas of pages
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
@param lru true=buf_pool.LRU; false=buf_pool.flush_list
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
@return last page number that can be flushed */
static page_id_t buf_flush_check_neighbors(const fil_space_t &space,
MDEV-23909 innodb_flush_neighbors=2 is treated like innodb_flush_neighbors=0 In MDEV-15053 (commit b1ab211dee599eabd9a5b886fafa3adea29ae041) we inadvertently removed a check whether innodb_flush_neighbors is 0, and thus started treating only the value 1 in a special way. buf_flush_check_neighbors(): Add the parameter contiguous, which can be set to skip the check for non-contiguous page number ranges. Reviewed by: Thirunarayanan Balathandayuthapani
2020-10-08 10:13:47 +02:00
page_id_t &id, bool contiguous,
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
bool lru)
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
{
MDEV-27016: Assertion 'id.page_no() < space.size' failed buf_flush_check_neighbors(): Relax a debug assertion that could fail for the very last page(s) of a ROW_FORMAT=COMPRESSED tables using a 1024-byte or 2048-byte page size. This assertion started to fail after commit d09426f9e60fd93296464ec9eb5f9d85566437d3 (MDEV-26537) modified the .ibd file extension to occur in steps of 4096 bytes.
2021-11-16 11:49:51 +01:00
ut_ad(id.page_no() < space.size +
(space.physical_size() == 2048 ? 1
: space.physical_size() == 1024 ? 3 : 0));
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
/* When flushed, dirty blocks are searched in neighborhoods of this
size, and flushed along with the original page. */
const ulint s= buf_pool.curr_size / 16;
const uint32_t read_ahead= buf_pool.read_ahead_area;
const uint32_t buf_flush_area= read_ahead > s
? static_cast<uint32_t>(s) : read_ahead;
page_id_t low= id - (id.page_no() % buf_flush_area);
page_id_t high= low + buf_flush_area;
MDEV-23855: Remove fil_system.LRU and reduce fil_system.mutex contention Also fixes MDEV-23929: innodb_flush_neighbors is not being ignored for system tablespace on SSD When the maximum configured number of file is exceeded, InnoDB will close data files. We used to maintain a fil_system.LRU list and a counter fil_node_t::n_pending to achieve this, at the huge cost of multiple fil_system.mutex operations per I/O operation. fil_node_open_file_low(): Implement a FIFO replacement policy: The last opened file will be moved to the end of fil_system.space_list, and files will be closed from the start of the list. However, we will not move tablespaces in fil_system.space_list while i_s_tablespaces_encryption_fill_table() is executing (producing output for INFORMATION_SCHEMA.INNODB_TABLESPACES_ENCRYPTION) because it may cause information of some tablespaces to go missing. We also avoid this in mariabackup --backup because datafiles_iter_next() assumes that the ordering is not changed. IORequest: Fold more parameters to IORequest::type. fil_space_t::io(): Replaces fil_io(). fil_space_t::flush(): Replaces fil_flush(). OS_AIO_IBUF: Remove. We will always issue synchronous reads of the change buffer pages in buf_read_page_low(). We will always ignore some errors for background reads. This should reduce fil_system.mutex contention a little. fil_node_t::complete_write(): Replaces fil_node_t::complete_io(). On both read and write completion, fil_space_t::release_for_io() will have to be called. fil_space_t::io(): Do not acquire fil_system.mutex in the normal code path. xb_delta_open_matching_space(): Do not try to open the system tablespace which was already opened. This fixes a file sharing violation in mariabackup --prepare --incremental. Reviewed by: Vladislav Vaintroub
2020-10-26 14:59:30 +01:00
high.set_page_no(std::min(high.page_no(), space.last_page_number()));
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
MDEV-23909 innodb_flush_neighbors=2 is treated like innodb_flush_neighbors=0 In MDEV-15053 (commit b1ab211dee599eabd9a5b886fafa3adea29ae041) we inadvertently removed a check whether innodb_flush_neighbors is 0, and thus started treating only the value 1 in a special way. buf_flush_check_neighbors(): Add the parameter contiguous, which can be set to skip the check for non-contiguous page number ranges. Reviewed by: Thirunarayanan Balathandayuthapani
2020-10-08 10:13:47 +02:00
if (!contiguous)
{
high= std::max(id + 1, high);
id= low;
return high;
}
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
/* Determine the contiguous dirty area around id. */
const ulint id_fold= id.fold();
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
mysql_mutex_lock(&buf_pool.mutex);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
if (id > low)
{
ulint fold= id_fold;
for (page_id_t i= id - 1;; --i)
{
fold--;
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
if (!buf_flush_check_neighbor(i, fold, lru))
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
{
low= i + 1;
break;
}
if (i == low)
break;
}
}
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
page_id_t i= id;
id= low;
ulint fold= id_fold;
while (++i < high)
{
++fold;
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
if (!buf_flush_check_neighbor(i, fold, lru))
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
break;
}
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
mysql_mutex_unlock(&buf_pool.mutex);
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
return i;
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
}
Cleanup: Make fil_space_t::freed_ranges private fil_space_t::is_freed(): Check if a page is in freed_ranges. fil_space_t::flush_freed(): Replaces buf_flush_freed_pages().
2022-06-06 10:55:29 +02:00
MY_ATTRIBUTE((warn_unused_result))
/** Apply freed_ranges to the file.
@param writable whether the file is writable
MDEV-27985 buf_flush_freed_pages() causes InnoDB to hang buf_flush_freed_pages(): Assert that neither buf_pool.mutex nor buf_pool.flush_list_mutex are held. Simplify the loops. Return the tablespace and the number of pages written or punched. buf_flush_LRU_list_batch(), buf_do_flush_list_batch(): Release buf_pool.mutex before invoking buf_flush_space(). buf_flush_list_space(): Acquire the mutexes only after invoking buf_flush_freed_pages(). Reviewed by: Thirunarayanan Balathandayuthapani
2022-03-15 13:44:22 +01:00
@return number of pages written or hole-punched */
Cleanup: Make fil_space_t::freed_ranges private fil_space_t::is_freed(): Check if a page is in freed_ranges. fil_space_t::flush_freed(): Replaces buf_flush_freed_pages().
2022-06-06 10:55:29 +02:00
uint32_t fil_space_t::flush_freed(bool writable)
MDEV-8139 Fix Scrubbing fil_space_t::freed_ranges: Store ranges of freed page numbers. fil_space_t::last_freed_lsn: Store the most recent LSN of freeing a page. fil_space_t::freed_mutex: Protects freed_ranges, last_freed_lsn. fil_space_create(): Initialize the freed_range mutex. fil_space_free_low(): Frees the freed_range mutex. range_set: Ranges of page numbers. buf_page_create(): Removes the page from freed_ranges when page is being reused. btr_free_root(): Remove the PAGE_INDEX_ID invalidation. Because btr_free_root() and dict_drop_index_tree() are executed in the same atomic mini-transaction, there is no need to invalidate the root page. buf_release_freed_page(): Split from buf_flush_freed_page(). Skip any I/O buf_flush_freed_pages(): Get the freed ranges from tablespace and Write punch-hole or zeroes of the freed ranges. buf_flush_try_neighbors(): Handles the flushing of freed ranges. mtr_t::freed_pages: Variable to store the list of freed pages. mtr_t::add_freed_pages(): To add freed pages. mtr_t::clear_freed_pages(): To clear the freed pages. mtr_t::m_freed_in_system_tablespace: Variable to indicate whether page has been freed in system tablespace. mtr_t::m_trim_pages: Variable to indicate whether the space has been trimmed. mtr_t::commit(): Add the freed page and update the last freed lsn in the tablespace and clear the tablespace freed range if space is trimmed. file_name_t::freed_pages: Store the freed pages during recovery. file_name_t::add_freed_page(), file_name_t::remove_freed_page(): To add and remove freed page during recovery. store_freed_or_init_rec(): Store or remove the freed pages while encountering FREE_PAGE or INIT_PAGE redo log record. recv_init_crash_recovery_spaces(): Add the freed page encountered during recovery to respective tablespace.
2020-06-11 19:22:47 +02:00
{
Cleanup: Make fil_space_t::freed_ranges private fil_space_t::is_freed(): Check if a page is in freed_ranges. fil_space_t::flush_freed(): Replaces buf_flush_freed_pages().
2022-06-06 10:55:29 +02:00
const bool punch_hole= chain.start->punch_hole == 1;
MDEV-26029: Sparse files are inefficient on thinly provisioned storage The MariaDB implementation of page_compressed tables for InnoDB used sparse files. In the worst case, in the data file, every data page will consist of some data followed by a hole. This may be extremely inefficient in some file systems. If the underlying storage device is thinly provisioned (can compress data on the fly), it would be good to write regular files (with sequences of NUL bytes at the end of each page_compressed block) and let the storage device take care of compressing the data. For reads, sparse file regions and regions containing NUL bytes will be indistinguishable. my_test_if_disable_punch_hole(): A new predicate for detecting thinly provisioned storage. (Not implemented yet.) innodb_atomic_writes: Correct the comment. buf_flush_page(): Support all values of fil_node_t::punch_hole. On a thinly provisioned storage device, we will always write NUL-padded innodb_page_size bytes also for page_compressed tables. buf_flush_freed_pages(): Remove a redundant condition. fil_space_t::atomic_write_supported: Remove. (This was duplicating fil_node_t::atomic_write.) fil_space_t::punch_hole: Remove. (Duplicated fil_node_t::punch_hole.) fil_node_t: Remove magic_n, and consolidate flags into bitfields. For punch_hole we introduce a third value that indicates a thinly provisioned storage device. fil_node_t::find_metadata(): Detect all attributes of the file.
2021-06-28 12:42:43 +02:00
if (!punch_hole && !srv_immediate_scrub_data_uncompressed)
MDEV-27985 buf_flush_freed_pages() causes InnoDB to hang buf_flush_freed_pages(): Assert that neither buf_pool.mutex nor buf_pool.flush_list_mutex are held. Simplify the loops. Return the tablespace and the number of pages written or punched. buf_flush_LRU_list_batch(), buf_do_flush_list_batch(): Release buf_pool.mutex before invoking buf_flush_space(). buf_flush_list_space(): Acquire the mutexes only after invoking buf_flush_freed_pages(). Reviewed by: Thirunarayanan Balathandayuthapani
2022-03-15 13:44:22 +01:00
return 0;
mysql_mutex_assert_not_owner(&buf_pool.flush_list_mutex);
mysql_mutex_assert_not_owner(&buf_pool.mutex);
MDEV-8139 Fix Scrubbing fil_space_t::freed_ranges: Store ranges of freed page numbers. fil_space_t::last_freed_lsn: Store the most recent LSN of freeing a page. fil_space_t::freed_mutex: Protects freed_ranges, last_freed_lsn. fil_space_create(): Initialize the freed_range mutex. fil_space_free_low(): Frees the freed_range mutex. range_set: Ranges of page numbers. buf_page_create(): Removes the page from freed_ranges when page is being reused. btr_free_root(): Remove the PAGE_INDEX_ID invalidation. Because btr_free_root() and dict_drop_index_tree() are executed in the same atomic mini-transaction, there is no need to invalidate the root page. buf_release_freed_page(): Split from buf_flush_freed_page(). Skip any I/O buf_flush_freed_pages(): Get the freed ranges from tablespace and Write punch-hole or zeroes of the freed ranges. buf_flush_try_neighbors(): Handles the flushing of freed ranges. mtr_t::freed_pages: Variable to store the list of freed pages. mtr_t::add_freed_pages(): To add freed pages. mtr_t::clear_freed_pages(): To clear the freed pages. mtr_t::m_freed_in_system_tablespace: Variable to indicate whether page has been freed in system tablespace. mtr_t::m_trim_pages: Variable to indicate whether the space has been trimmed. mtr_t::commit(): Add the freed page and update the last freed lsn in the tablespace and clear the tablespace freed range if space is trimmed. file_name_t::freed_pages: Store the freed pages during recovery. file_name_t::add_freed_page(), file_name_t::remove_freed_page(): To add and remove freed page during recovery. store_freed_or_init_rec(): Store or remove the freed pages while encountering FREE_PAGE or INIT_PAGE redo log record. recv_init_crash_recovery_spaces(): Add the freed page encountered during recovery to respective tablespace.
2020-06-11 19:22:47 +02:00
Cleanup: Make fil_space_t::freed_ranges private fil_space_t::is_freed(): Check if a page is in freed_ranges. fil_space_t::flush_freed(): Replaces buf_flush_freed_pages().
2022-06-06 10:55:29 +02:00
freed_range_mutex.lock();
if (freed_ranges.empty() || log_sys.get_flushed_lsn() < get_last_freed_lsn())
MDEV-8139 Fix Scrubbing fil_space_t::freed_ranges: Store ranges of freed page numbers. fil_space_t::last_freed_lsn: Store the most recent LSN of freeing a page. fil_space_t::freed_mutex: Protects freed_ranges, last_freed_lsn. fil_space_create(): Initialize the freed_range mutex. fil_space_free_low(): Frees the freed_range mutex. range_set: Ranges of page numbers. buf_page_create(): Removes the page from freed_ranges when page is being reused. btr_free_root(): Remove the PAGE_INDEX_ID invalidation. Because btr_free_root() and dict_drop_index_tree() are executed in the same atomic mini-transaction, there is no need to invalidate the root page. buf_release_freed_page(): Split from buf_flush_freed_page(). Skip any I/O buf_flush_freed_pages(): Get the freed ranges from tablespace and Write punch-hole or zeroes of the freed ranges. buf_flush_try_neighbors(): Handles the flushing of freed ranges. mtr_t::freed_pages: Variable to store the list of freed pages. mtr_t::add_freed_pages(): To add freed pages. mtr_t::clear_freed_pages(): To clear the freed pages. mtr_t::m_freed_in_system_tablespace: Variable to indicate whether page has been freed in system tablespace. mtr_t::m_trim_pages: Variable to indicate whether the space has been trimmed. mtr_t::commit(): Add the freed page and update the last freed lsn in the tablespace and clear the tablespace freed range if space is trimmed. file_name_t::freed_pages: Store the freed pages during recovery. file_name_t::add_freed_page(), file_name_t::remove_freed_page(): To add and remove freed page during recovery. store_freed_or_init_rec(): Store or remove the freed pages while encountering FREE_PAGE or INIT_PAGE redo log record. recv_init_crash_recovery_spaces(): Add the freed page encountered during recovery to respective tablespace.
2020-06-11 19:22:47 +02:00
{
Cleanup: Make fil_space_t::freed_ranges private fil_space_t::is_freed(): Check if a page is in freed_ranges. fil_space_t::flush_freed(): Replaces buf_flush_freed_pages().
2022-06-06 10:55:29 +02:00
freed_range_mutex.unlock();
MDEV-27985 buf_flush_freed_pages() causes InnoDB to hang buf_flush_freed_pages(): Assert that neither buf_pool.mutex nor buf_pool.flush_list_mutex are held. Simplify the loops. Return the tablespace and the number of pages written or punched. buf_flush_LRU_list_batch(), buf_do_flush_list_batch(): Release buf_pool.mutex before invoking buf_flush_space(). buf_flush_list_space(): Acquire the mutexes only after invoking buf_flush_freed_pages(). Reviewed by: Thirunarayanan Balathandayuthapani
2022-03-15 13:44:22 +01:00
return 0;
MDEV-8139 Fix Scrubbing fil_space_t::freed_ranges: Store ranges of freed page numbers. fil_space_t::last_freed_lsn: Store the most recent LSN of freeing a page. fil_space_t::freed_mutex: Protects freed_ranges, last_freed_lsn. fil_space_create(): Initialize the freed_range mutex. fil_space_free_low(): Frees the freed_range mutex. range_set: Ranges of page numbers. buf_page_create(): Removes the page from freed_ranges when page is being reused. btr_free_root(): Remove the PAGE_INDEX_ID invalidation. Because btr_free_root() and dict_drop_index_tree() are executed in the same atomic mini-transaction, there is no need to invalidate the root page. buf_release_freed_page(): Split from buf_flush_freed_page(). Skip any I/O buf_flush_freed_pages(): Get the freed ranges from tablespace and Write punch-hole or zeroes of the freed ranges. buf_flush_try_neighbors(): Handles the flushing of freed ranges. mtr_t::freed_pages: Variable to store the list of freed pages. mtr_t::add_freed_pages(): To add freed pages. mtr_t::clear_freed_pages(): To clear the freed pages. mtr_t::m_freed_in_system_tablespace: Variable to indicate whether page has been freed in system tablespace. mtr_t::m_trim_pages: Variable to indicate whether the space has been trimmed. mtr_t::commit(): Add the freed page and update the last freed lsn in the tablespace and clear the tablespace freed range if space is trimmed. file_name_t::freed_pages: Store the freed pages during recovery. file_name_t::add_freed_page(), file_name_t::remove_freed_page(): To add and remove freed page during recovery. store_freed_or_init_rec(): Store or remove the freed pages while encountering FREE_PAGE or INIT_PAGE redo log record. recv_init_crash_recovery_spaces(): Add the freed page encountered during recovery to respective tablespace.
2020-06-11 19:22:47 +02:00
}
Cleanup: Make fil_space_t::freed_ranges private fil_space_t::is_freed(): Check if a page is in freed_ranges. fil_space_t::flush_freed(): Replaces buf_flush_freed_pages().
2022-06-06 10:55:29 +02:00
const unsigned physical{physical_size()};
MDEV-27985 buf_flush_freed_pages() causes InnoDB to hang buf_flush_freed_pages(): Assert that neither buf_pool.mutex nor buf_pool.flush_list_mutex are held. Simplify the loops. Return the tablespace and the number of pages written or punched. buf_flush_LRU_list_batch(), buf_do_flush_list_batch(): Release buf_pool.mutex before invoking buf_flush_space(). buf_flush_list_space(): Acquire the mutexes only after invoking buf_flush_freed_pages(). Reviewed by: Thirunarayanan Balathandayuthapani
2022-03-15 13:44:22 +01:00
Cleanup: Make fil_space_t::freed_ranges private fil_space_t::is_freed(): Check if a page is in freed_ranges. fil_space_t::flush_freed(): Replaces buf_flush_freed_pages().
2022-06-06 10:55:29 +02:00
range_set freed= std::move(freed_ranges);
MDEV-27985 buf_flush_freed_pages() causes InnoDB to hang buf_flush_freed_pages(): Assert that neither buf_pool.mutex nor buf_pool.flush_list_mutex are held. Simplify the loops. Return the tablespace and the number of pages written or punched. buf_flush_LRU_list_batch(), buf_do_flush_list_batch(): Release buf_pool.mutex before invoking buf_flush_space(). buf_flush_list_space(): Acquire the mutexes only after invoking buf_flush_freed_pages(). Reviewed by: Thirunarayanan Balathandayuthapani
2022-03-15 13:44:22 +01:00
uint32_t written= 0;
MDEV-8139 Fix Scrubbing fil_space_t::freed_ranges: Store ranges of freed page numbers. fil_space_t::last_freed_lsn: Store the most recent LSN of freeing a page. fil_space_t::freed_mutex: Protects freed_ranges, last_freed_lsn. fil_space_create(): Initialize the freed_range mutex. fil_space_free_low(): Frees the freed_range mutex. range_set: Ranges of page numbers. buf_page_create(): Removes the page from freed_ranges when page is being reused. btr_free_root(): Remove the PAGE_INDEX_ID invalidation. Because btr_free_root() and dict_drop_index_tree() are executed in the same atomic mini-transaction, there is no need to invalidate the root page. buf_release_freed_page(): Split from buf_flush_freed_page(). Skip any I/O buf_flush_freed_pages(): Get the freed ranges from tablespace and Write punch-hole or zeroes of the freed ranges. buf_flush_try_neighbors(): Handles the flushing of freed ranges. mtr_t::freed_pages: Variable to store the list of freed pages. mtr_t::add_freed_pages(): To add freed pages. mtr_t::clear_freed_pages(): To clear the freed pages. mtr_t::m_freed_in_system_tablespace: Variable to indicate whether page has been freed in system tablespace. mtr_t::m_trim_pages: Variable to indicate whether the space has been trimmed. mtr_t::commit(): Add the freed page and update the last freed lsn in the tablespace and clear the tablespace freed range if space is trimmed. file_name_t::freed_pages: Store the freed pages during recovery. file_name_t::add_freed_page(), file_name_t::remove_freed_page(): To add and remove freed page during recovery. store_freed_or_init_rec(): Store or remove the freed pages while encountering FREE_PAGE or INIT_PAGE redo log record. recv_init_crash_recovery_spaces(): Add the freed page encountered during recovery to respective tablespace.
2020-06-11 19:22:47 +02:00
MDEV-27985 buf_flush_freed_pages() causes InnoDB to hang buf_flush_freed_pages(): Assert that neither buf_pool.mutex nor buf_pool.flush_list_mutex are held. Simplify the loops. Return the tablespace and the number of pages written or punched. buf_flush_LRU_list_batch(), buf_do_flush_list_batch(): Release buf_pool.mutex before invoking buf_flush_space(). buf_flush_list_space(): Acquire the mutexes only after invoking buf_flush_freed_pages(). Reviewed by: Thirunarayanan Balathandayuthapani
2022-03-15 13:44:22 +01:00
if (!writable);
else if (punch_hole)
MDEV-8139 Fix Scrubbing fil_space_t::freed_ranges: Store ranges of freed page numbers. fil_space_t::last_freed_lsn: Store the most recent LSN of freeing a page. fil_space_t::freed_mutex: Protects freed_ranges, last_freed_lsn. fil_space_create(): Initialize the freed_range mutex. fil_space_free_low(): Frees the freed_range mutex. range_set: Ranges of page numbers. buf_page_create(): Removes the page from freed_ranges when page is being reused. btr_free_root(): Remove the PAGE_INDEX_ID invalidation. Because btr_free_root() and dict_drop_index_tree() are executed in the same atomic mini-transaction, there is no need to invalidate the root page. buf_release_freed_page(): Split from buf_flush_freed_page(). Skip any I/O buf_flush_freed_pages(): Get the freed ranges from tablespace and Write punch-hole or zeroes of the freed ranges. buf_flush_try_neighbors(): Handles the flushing of freed ranges. mtr_t::freed_pages: Variable to store the list of freed pages. mtr_t::add_freed_pages(): To add freed pages. mtr_t::clear_freed_pages(): To clear the freed pages. mtr_t::m_freed_in_system_tablespace: Variable to indicate whether page has been freed in system tablespace. mtr_t::m_trim_pages: Variable to indicate whether the space has been trimmed. mtr_t::commit(): Add the freed page and update the last freed lsn in the tablespace and clear the tablespace freed range if space is trimmed. file_name_t::freed_pages: Store the freed pages during recovery. file_name_t::add_freed_page(), file_name_t::remove_freed_page(): To add and remove freed page during recovery. store_freed_or_init_rec(): Store or remove the freed pages while encountering FREE_PAGE or INIT_PAGE redo log record. recv_init_crash_recovery_spaces(): Add the freed page encountered during recovery to respective tablespace.
2020-06-11 19:22:47 +02:00
{
Cleanup: Make fil_space_t::freed_ranges private fil_space_t::is_freed(): Check if a page is in freed_ranges. fil_space_t::flush_freed(): Replaces buf_flush_freed_pages().
2022-06-06 10:55:29 +02:00
for (const auto &range : freed)
MDEV-8139 Fix Scrubbing fil_space_t::freed_ranges: Store ranges of freed page numbers. fil_space_t::last_freed_lsn: Store the most recent LSN of freeing a page. fil_space_t::freed_mutex: Protects freed_ranges, last_freed_lsn. fil_space_create(): Initialize the freed_range mutex. fil_space_free_low(): Frees the freed_range mutex. range_set: Ranges of page numbers. buf_page_create(): Removes the page from freed_ranges when page is being reused. btr_free_root(): Remove the PAGE_INDEX_ID invalidation. Because btr_free_root() and dict_drop_index_tree() are executed in the same atomic mini-transaction, there is no need to invalidate the root page. buf_release_freed_page(): Split from buf_flush_freed_page(). Skip any I/O buf_flush_freed_pages(): Get the freed ranges from tablespace and Write punch-hole or zeroes of the freed ranges. buf_flush_try_neighbors(): Handles the flushing of freed ranges. mtr_t::freed_pages: Variable to store the list of freed pages. mtr_t::add_freed_pages(): To add freed pages. mtr_t::clear_freed_pages(): To clear the freed pages. mtr_t::m_freed_in_system_tablespace: Variable to indicate whether page has been freed in system tablespace. mtr_t::m_trim_pages: Variable to indicate whether the space has been trimmed. mtr_t::commit(): Add the freed page and update the last freed lsn in the tablespace and clear the tablespace freed range if space is trimmed. file_name_t::freed_pages: Store the freed pages during recovery. file_name_t::add_freed_page(), file_name_t::remove_freed_page(): To add and remove freed page during recovery. store_freed_or_init_rec(): Store or remove the freed pages while encountering FREE_PAGE or INIT_PAGE redo log record. recv_init_crash_recovery_spaces(): Add the freed page encountered during recovery to respective tablespace.
2020-06-11 19:22:47 +02:00
{
MDEV-27985 buf_flush_freed_pages() causes InnoDB to hang buf_flush_freed_pages(): Assert that neither buf_pool.mutex nor buf_pool.flush_list_mutex are held. Simplify the loops. Return the tablespace and the number of pages written or punched. buf_flush_LRU_list_batch(), buf_do_flush_list_batch(): Release buf_pool.mutex before invoking buf_flush_space(). buf_flush_list_space(): Acquire the mutexes only after invoking buf_flush_freed_pages(). Reviewed by: Thirunarayanan Balathandayuthapani
2022-03-15 13:44:22 +01:00
written+= range.last - range.first + 1;
Cleanup: Make fil_space_t::freed_ranges private fil_space_t::is_freed(): Check if a page is in freed_ranges. fil_space_t::flush_freed(): Replaces buf_flush_freed_pages().
2022-06-06 10:55:29 +02:00
reacquire();
io(IORequest(IORequest::PUNCH_RANGE),
os_offset_t{range.first} * physical,
(range.last - range.first + 1) * physical, nullptr);
MDEV-8139 Fix Scrubbing fil_space_t::freed_ranges: Store ranges of freed page numbers. fil_space_t::last_freed_lsn: Store the most recent LSN of freeing a page. fil_space_t::freed_mutex: Protects freed_ranges, last_freed_lsn. fil_space_create(): Initialize the freed_range mutex. fil_space_free_low(): Frees the freed_range mutex. range_set: Ranges of page numbers. buf_page_create(): Removes the page from freed_ranges when page is being reused. btr_free_root(): Remove the PAGE_INDEX_ID invalidation. Because btr_free_root() and dict_drop_index_tree() are executed in the same atomic mini-transaction, there is no need to invalidate the root page. buf_release_freed_page(): Split from buf_flush_freed_page(). Skip any I/O buf_flush_freed_pages(): Get the freed ranges from tablespace and Write punch-hole or zeroes of the freed ranges. buf_flush_try_neighbors(): Handles the flushing of freed ranges. mtr_t::freed_pages: Variable to store the list of freed pages. mtr_t::add_freed_pages(): To add freed pages. mtr_t::clear_freed_pages(): To clear the freed pages. mtr_t::m_freed_in_system_tablespace: Variable to indicate whether page has been freed in system tablespace. mtr_t::m_trim_pages: Variable to indicate whether the space has been trimmed. mtr_t::commit(): Add the freed page and update the last freed lsn in the tablespace and clear the tablespace freed range if space is trimmed. file_name_t::freed_pages: Store the freed pages during recovery. file_name_t::add_freed_page(), file_name_t::remove_freed_page(): To add and remove freed page during recovery. store_freed_or_init_rec(): Store or remove the freed pages while encountering FREE_PAGE or INIT_PAGE redo log record. recv_init_crash_recovery_spaces(): Add the freed page encountered during recovery to respective tablespace.
2020-06-11 19:22:47 +02:00
}
MDEV-27985 buf_flush_freed_pages() causes InnoDB to hang buf_flush_freed_pages(): Assert that neither buf_pool.mutex nor buf_pool.flush_list_mutex are held. Simplify the loops. Return the tablespace and the number of pages written or punched. buf_flush_LRU_list_batch(), buf_do_flush_list_batch(): Release buf_pool.mutex before invoking buf_flush_space(). buf_flush_list_space(): Acquire the mutexes only after invoking buf_flush_freed_pages(). Reviewed by: Thirunarayanan Balathandayuthapani
2022-03-15 13:44:22 +01:00
}
else
{
Cleanup: Make fil_space_t::freed_ranges private fil_space_t::is_freed(): Check if a page is in freed_ranges. fil_space_t::flush_freed(): Replaces buf_flush_freed_pages().
2022-06-06 10:55:29 +02:00
for (const auto &range : freed)
MDEV-8139 Fix Scrubbing fil_space_t::freed_ranges: Store ranges of freed page numbers. fil_space_t::last_freed_lsn: Store the most recent LSN of freeing a page. fil_space_t::freed_mutex: Protects freed_ranges, last_freed_lsn. fil_space_create(): Initialize the freed_range mutex. fil_space_free_low(): Frees the freed_range mutex. range_set: Ranges of page numbers. buf_page_create(): Removes the page from freed_ranges when page is being reused. btr_free_root(): Remove the PAGE_INDEX_ID invalidation. Because btr_free_root() and dict_drop_index_tree() are executed in the same atomic mini-transaction, there is no need to invalidate the root page. buf_release_freed_page(): Split from buf_flush_freed_page(). Skip any I/O buf_flush_freed_pages(): Get the freed ranges from tablespace and Write punch-hole or zeroes of the freed ranges. buf_flush_try_neighbors(): Handles the flushing of freed ranges. mtr_t::freed_pages: Variable to store the list of freed pages. mtr_t::add_freed_pages(): To add freed pages. mtr_t::clear_freed_pages(): To clear the freed pages. mtr_t::m_freed_in_system_tablespace: Variable to indicate whether page has been freed in system tablespace. mtr_t::m_trim_pages: Variable to indicate whether the space has been trimmed. mtr_t::commit(): Add the freed page and update the last freed lsn in the tablespace and clear the tablespace freed range if space is trimmed. file_name_t::freed_pages: Store the freed pages during recovery. file_name_t::add_freed_page(), file_name_t::remove_freed_page(): To add and remove freed page during recovery. store_freed_or_init_rec(): Store or remove the freed pages while encountering FREE_PAGE or INIT_PAGE redo log record. recv_init_crash_recovery_spaces(): Add the freed page encountered during recovery to respective tablespace.
2020-06-11 19:22:47 +02:00
{
MDEV-27985 buf_flush_freed_pages() causes InnoDB to hang buf_flush_freed_pages(): Assert that neither buf_pool.mutex nor buf_pool.flush_list_mutex are held. Simplify the loops. Return the tablespace and the number of pages written or punched. buf_flush_LRU_list_batch(), buf_do_flush_list_batch(): Release buf_pool.mutex before invoking buf_flush_space(). buf_flush_list_space(): Acquire the mutexes only after invoking buf_flush_freed_pages(). Reviewed by: Thirunarayanan Balathandayuthapani
2022-03-15 13:44:22 +01:00
written+= range.last - range.first + 1;
MDEV-23855: Remove fil_system.LRU and reduce fil_system.mutex contention Also fixes MDEV-23929: innodb_flush_neighbors is not being ignored for system tablespace on SSD When the maximum configured number of file is exceeded, InnoDB will close data files. We used to maintain a fil_system.LRU list and a counter fil_node_t::n_pending to achieve this, at the huge cost of multiple fil_system.mutex operations per I/O operation. fil_node_open_file_low(): Implement a FIFO replacement policy: The last opened file will be moved to the end of fil_system.space_list, and files will be closed from the start of the list. However, we will not move tablespaces in fil_system.space_list while i_s_tablespaces_encryption_fill_table() is executing (producing output for INFORMATION_SCHEMA.INNODB_TABLESPACES_ENCRYPTION) because it may cause information of some tablespaces to go missing. We also avoid this in mariabackup --backup because datafiles_iter_next() assumes that the ordering is not changed. IORequest: Fold more parameters to IORequest::type. fil_space_t::io(): Replaces fil_io(). fil_space_t::flush(): Replaces fil_flush(). OS_AIO_IBUF: Remove. We will always issue synchronous reads of the change buffer pages in buf_read_page_low(). We will always ignore some errors for background reads. This should reduce fil_system.mutex contention a little. fil_node_t::complete_write(): Replaces fil_node_t::complete_io(). On both read and write completion, fil_space_t::release_for_io() will have to be called. fil_space_t::io(): Do not acquire fil_system.mutex in the normal code path. xb_delta_open_matching_space(): Do not try to open the system tablespace which was already opened. This fixes a file sharing violation in mariabackup --prepare --incremental. Reviewed by: Vladislav Vaintroub
2020-10-26 14:59:30 +01:00
for (os_offset_t i= range.first; i <= range.last; i++)
MDEV-8139 Fix Scrubbing fil_space_t::freed_ranges: Store ranges of freed page numbers. fil_space_t::last_freed_lsn: Store the most recent LSN of freeing a page. fil_space_t::freed_mutex: Protects freed_ranges, last_freed_lsn. fil_space_create(): Initialize the freed_range mutex. fil_space_free_low(): Frees the freed_range mutex. range_set: Ranges of page numbers. buf_page_create(): Removes the page from freed_ranges when page is being reused. btr_free_root(): Remove the PAGE_INDEX_ID invalidation. Because btr_free_root() and dict_drop_index_tree() are executed in the same atomic mini-transaction, there is no need to invalidate the root page. buf_release_freed_page(): Split from buf_flush_freed_page(). Skip any I/O buf_flush_freed_pages(): Get the freed ranges from tablespace and Write punch-hole or zeroes of the freed ranges. buf_flush_try_neighbors(): Handles the flushing of freed ranges. mtr_t::freed_pages: Variable to store the list of freed pages. mtr_t::add_freed_pages(): To add freed pages. mtr_t::clear_freed_pages(): To clear the freed pages. mtr_t::m_freed_in_system_tablespace: Variable to indicate whether page has been freed in system tablespace. mtr_t::m_trim_pages: Variable to indicate whether the space has been trimmed. mtr_t::commit(): Add the freed page and update the last freed lsn in the tablespace and clear the tablespace freed range if space is trimmed. file_name_t::freed_pages: Store the freed pages during recovery. file_name_t::add_freed_page(), file_name_t::remove_freed_page(): To add and remove freed page during recovery. store_freed_or_init_rec(): Store or remove the freed pages while encountering FREE_PAGE or INIT_PAGE redo log record. recv_init_crash_recovery_spaces(): Add the freed page encountered during recovery to respective tablespace.
2020-06-11 19:22:47 +02:00
{
Cleanup: Make fil_space_t::freed_ranges private fil_space_t::is_freed(): Check if a page is in freed_ranges. fil_space_t::flush_freed(): Replaces buf_flush_freed_pages().
2022-06-06 10:55:29 +02:00
reacquire();
io(IORequest(IORequest::WRITE_ASYNC), i * physical, physical,
const_cast<byte*>(field_ref_zero));
MDEV-8139 Fix Scrubbing fil_space_t::freed_ranges: Store ranges of freed page numbers. fil_space_t::last_freed_lsn: Store the most recent LSN of freeing a page. fil_space_t::freed_mutex: Protects freed_ranges, last_freed_lsn. fil_space_create(): Initialize the freed_range mutex. fil_space_free_low(): Frees the freed_range mutex. range_set: Ranges of page numbers. buf_page_create(): Removes the page from freed_ranges when page is being reused. btr_free_root(): Remove the PAGE_INDEX_ID invalidation. Because btr_free_root() and dict_drop_index_tree() are executed in the same atomic mini-transaction, there is no need to invalidate the root page. buf_release_freed_page(): Split from buf_flush_freed_page(). Skip any I/O buf_flush_freed_pages(): Get the freed ranges from tablespace and Write punch-hole or zeroes of the freed ranges. buf_flush_try_neighbors(): Handles the flushing of freed ranges. mtr_t::freed_pages: Variable to store the list of freed pages. mtr_t::add_freed_pages(): To add freed pages. mtr_t::clear_freed_pages(): To clear the freed pages. mtr_t::m_freed_in_system_tablespace: Variable to indicate whether page has been freed in system tablespace. mtr_t::m_trim_pages: Variable to indicate whether the space has been trimmed. mtr_t::commit(): Add the freed page and update the last freed lsn in the tablespace and clear the tablespace freed range if space is trimmed. file_name_t::freed_pages: Store the freed pages during recovery. file_name_t::add_freed_page(), file_name_t::remove_freed_page(): To add and remove freed page during recovery. store_freed_or_init_rec(): Store or remove the freed pages while encountering FREE_PAGE or INIT_PAGE redo log record. recv_init_crash_recovery_spaces(): Add the freed page encountered during recovery to respective tablespace.
2020-06-11 19:22:47 +02:00
}
}
}
MDEV-27985 buf_flush_freed_pages() causes InnoDB to hang buf_flush_freed_pages(): Assert that neither buf_pool.mutex nor buf_pool.flush_list_mutex are held. Simplify the loops. Return the tablespace and the number of pages written or punched. buf_flush_LRU_list_batch(), buf_do_flush_list_batch(): Release buf_pool.mutex before invoking buf_flush_space(). buf_flush_list_space(): Acquire the mutexes only after invoking buf_flush_freed_pages(). Reviewed by: Thirunarayanan Balathandayuthapani
2022-03-15 13:44:22 +01:00
Cleanup: Make fil_space_t::freed_ranges private fil_space_t::is_freed(): Check if a page is in freed_ranges. fil_space_t::flush_freed(): Replaces buf_flush_freed_pages().
2022-06-06 10:55:29 +02:00
freed_range_mutex.unlock();
MDEV-27985 buf_flush_freed_pages() causes InnoDB to hang buf_flush_freed_pages(): Assert that neither buf_pool.mutex nor buf_pool.flush_list_mutex are held. Simplify the loops. Return the tablespace and the number of pages written or punched. buf_flush_LRU_list_batch(), buf_do_flush_list_batch(): Release buf_pool.mutex before invoking buf_flush_space(). buf_flush_list_space(): Acquire the mutexes only after invoking buf_flush_freed_pages(). Reviewed by: Thirunarayanan Balathandayuthapani
2022-03-15 13:44:22 +01:00
return written;
MDEV-8139 Fix Scrubbing fil_space_t::freed_ranges: Store ranges of freed page numbers. fil_space_t::last_freed_lsn: Store the most recent LSN of freeing a page. fil_space_t::freed_mutex: Protects freed_ranges, last_freed_lsn. fil_space_create(): Initialize the freed_range mutex. fil_space_free_low(): Frees the freed_range mutex. range_set: Ranges of page numbers. buf_page_create(): Removes the page from freed_ranges when page is being reused. btr_free_root(): Remove the PAGE_INDEX_ID invalidation. Because btr_free_root() and dict_drop_index_tree() are executed in the same atomic mini-transaction, there is no need to invalidate the root page. buf_release_freed_page(): Split from buf_flush_freed_page(). Skip any I/O buf_flush_freed_pages(): Get the freed ranges from tablespace and Write punch-hole or zeroes of the freed ranges. buf_flush_try_neighbors(): Handles the flushing of freed ranges. mtr_t::freed_pages: Variable to store the list of freed pages. mtr_t::add_freed_pages(): To add freed pages. mtr_t::clear_freed_pages(): To clear the freed pages. mtr_t::m_freed_in_system_tablespace: Variable to indicate whether page has been freed in system tablespace. mtr_t::m_trim_pages: Variable to indicate whether the space has been trimmed. mtr_t::commit(): Add the freed page and update the last freed lsn in the tablespace and clear the tablespace freed range if space is trimmed. file_name_t::freed_pages: Store the freed pages during recovery. file_name_t::add_freed_page(), file_name_t::remove_freed_page(): To add and remove freed page during recovery. store_freed_or_init_rec(): Store or remove the freed pages while encountering FREE_PAGE or INIT_PAGE redo log record. recv_init_crash_recovery_spaces(): Add the freed page encountered during recovery to respective tablespace.
2020-06-11 19:22:47 +02:00
}
/** Flushes to disk all flushable pages within the flush area
and also write zeroes or punch the hole for the freed ranges of pages.
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
@param space tablespace
@param page_id page identifier
@param contiguous whether to consider contiguous areas of pages
@param lru true=buf_pool.LRU; false=buf_pool.flush_list
@param n_flushed number of pages flushed so far in this batch
@param n_to_flush maximum number of pages we are allowed to flush
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
@return number of pages flushed */
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
static ulint buf_flush_try_neighbors(fil_space_t *space,
const page_id_t page_id,
bool contiguous, bool lru,
ulint n_flushed, ulint n_to_flush)
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
{
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
ut_ad(space->id == page_id.space());
MDEV-23909 innodb_flush_neighbors=2 is treated like innodb_flush_neighbors=0 In MDEV-15053 (commit b1ab211dee599eabd9a5b886fafa3adea29ae041) we inadvertently removed a check whether innodb_flush_neighbors is 0, and thus started treating only the value 1 in a special way. buf_flush_check_neighbors(): Add the parameter contiguous, which can be set to skip the check for non-contiguous page number ranges. Reviewed by: Thirunarayanan Balathandayuthapani
2020-10-08 10:13:47 +02:00
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
ulint count= 0;
page_id_t id= page_id;
page_id_t high= buf_flush_check_neighbors(*space, id, contiguous, lru);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
ut_ad(page_id >= id);
ut_ad(page_id < high);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-23855: Remove fil_system.LRU and reduce fil_system.mutex contention Also fixes MDEV-23929: innodb_flush_neighbors is not being ignored for system tablespace on SSD When the maximum configured number of file is exceeded, InnoDB will close data files. We used to maintain a fil_system.LRU list and a counter fil_node_t::n_pending to achieve this, at the huge cost of multiple fil_system.mutex operations per I/O operation. fil_node_open_file_low(): Implement a FIFO replacement policy: The last opened file will be moved to the end of fil_system.space_list, and files will be closed from the start of the list. However, we will not move tablespaces in fil_system.space_list while i_s_tablespaces_encryption_fill_table() is executing (producing output for INFORMATION_SCHEMA.INNODB_TABLESPACES_ENCRYPTION) because it may cause information of some tablespaces to go missing. We also avoid this in mariabackup --backup because datafiles_iter_next() assumes that the ordering is not changed. IORequest: Fold more parameters to IORequest::type. fil_space_t::io(): Replaces fil_io(). fil_space_t::flush(): Replaces fil_flush(). OS_AIO_IBUF: Remove. We will always issue synchronous reads of the change buffer pages in buf_read_page_low(). We will always ignore some errors for background reads. This should reduce fil_system.mutex contention a little. fil_node_t::complete_write(): Replaces fil_node_t::complete_io(). On both read and write completion, fil_space_t::release_for_io() will have to be called. fil_space_t::io(): Do not acquire fil_system.mutex in the normal code path. xb_delta_open_matching_space(): Do not try to open the system tablespace which was already opened. This fixes a file sharing violation in mariabackup --prepare --incremental. Reviewed by: Vladislav Vaintroub
2020-10-26 14:59:30 +01:00
for (ulint id_fold= id.fold(); id < high && !space->is_stopping();
++id, ++id_fold)
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
{
if (count + n_flushed >= n_to_flush)
{
if (id > page_id)
break;
/* If the page whose neighbors we are flushing has not been
flushed yet, we must flush the page that we selected originally. */
id= page_id;
id_fold= id.fold();
}
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-26826 Duplicated computations of buf_pool.page_hash addresses Since commit bd5a6403cace36c6ed428cde62e35adcd3f7e7d0 (MDEV-26033) we can actually calculate the buf_pool.page_hash cell and latch addresses while not holding buf_pool.mutex. buf_page_alloc_descriptor(): Remove the MEM_UNDEFINED. We now expect buf_page_t::hash to be zero-initialized. buf_pool_t::hash_chain: Dedicated data type for buf_pool.page_hash.array. buf_LRU_free_one_page(): Merged to the only caller buf_pool_t::corrupted_evict().
2021-10-22 11:33:37 +02:00
const buf_pool_t::hash_chain &chain= buf_pool.page_hash.cell_get(id_fold);
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
mysql_mutex_lock(&buf_pool.mutex);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-26826 Duplicated computations of buf_pool.page_hash addresses Since commit bd5a6403cace36c6ed428cde62e35adcd3f7e7d0 (MDEV-26033) we can actually calculate the buf_pool.page_hash cell and latch addresses while not holding buf_pool.mutex. buf_page_alloc_descriptor(): Remove the MEM_UNDEFINED. We now expect buf_page_t::hash to be zero-initialized. buf_pool_t::hash_chain: Dedicated data type for buf_pool.page_hash.array. buf_LRU_free_one_page(): Merged to the only caller buf_pool_t::corrupted_evict().
2021-10-22 11:33:37 +02:00
if (buf_page_t *bpage= buf_pool.page_hash.get(id, chain))
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
{
ut_ad(bpage->in_file());
/* We avoid flushing 'non-old' blocks in an LRU flush,
because the flushed blocks are soon freed */
if (!lru || id == page_id || bpage->is_old())
{
MDEV-24054 Assertion in_LRU_list failed in buf_flush_try_neighbors() buf_flush_try_neighbors(): Before invoking buf_page_t::ready_for_flush(), check that the freshly looked up buf_pool.page_hash entry actually is a buffer page and not a buf_pool.watch[] sentinel for purge buffering. This race condition was introduced in MDEV-15053 (commit b1ab211dee599eabd9a5b886fafa3adea29ae041). It is rather hard to hit this bug, because buf_flush_check_neighbors() already checked the condition. The problem exists if buf_pool.watch_set() was invoked for a page in the range after the check in buf_flush_check_neighbor() had been finished.
2020-10-30 18:06:50 +01:00
if (!buf_pool.watch_is_sentinel(*bpage) &&
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
bpage->oldest_modification() > 1 && bpage->ready_for_flush() &&
bpage->flush(lru, space))
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
{
++count;
continue;
}
}
}
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
mysql_mutex_unlock(&buf_pool.mutex);
}
MDEV-17380 innodb_flush_neighbors=ON should be ignored on SSD For tablespaces that do not reside on spinning storage, it does not make sense to attempt to write nearby pages when writing out dirty pages from the InnoDB buffer pool. It is actually detrimental to performance and to the life span of flash ROM storage. With this change, MariaDB will detect whether an InnoDB file resides on solid-state storage. The detection has been implemented for Linux and Microsoft Windows. For other systems, we will err on the safe side and assume that files reside on SSD. As part of this change, we will reduce the number of fstat() calls when opening data files on POSIX systems and slightly clean up some file I/O code. FIXME: os_is_sparse_file_supported() on POSIX works in a destructive manner. Thus, we can only invoke it when creating files, not when opening them. For diagnostics, we introduce the column ON_SSD to the table INFORMATION_SCHEMA.INNODB_TABLESPACES_SCRUBBING. The table INNODB_SYS_TABLESPACES might seem more appropriate, but its purpose is to reflect the contents of the InnoDB system table SYS_TABLESPACES, which we would like to remove at some point. On Microsoft Windows, querying StorageDeviceSeekPenaltyProperty sometimes returns ERROR_GEN_FAILURE instead of ERROR_INVALID_FUNCTION or ERROR_NOT_SUPPORTED. We will silently ignore also this error, and assume that the file does not reside on SSD. On Linux, the detection will be based on the files /sys/block/*/queue/rotational and /sys/block/*/dev. Especially for USB storage, it is possible that /sys/block/*/queue/rotational will wrongly report 1 instead of 0. fil_node_t::on_ssd: Whether the InnoDB data file resides on solid-state storage. fil_system_t::ssd: Collection of Linux block devices that reside on non-rotational storage. fil_system_t::create(): Detect ssd on Linux based on the contents of /sys/block/*/queue/rotational and /sys/block/*/dev. fil_system_t::is_ssd(dev_t): Determine if a Linux block device is non-rotational. Partitions will be identified with the containing block device by assuming that the least significant 4 bits of the minor number identify a partition, and that the "partition number" of the entire device is 0.
2019-04-01 10:57:06 +02:00
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
if (auto n= count - 1)
{
MONITOR_INC_VALUE_CUMULATIVE(MONITOR_FLUSH_NEIGHBOR_TOTAL_PAGE,
MONITOR_FLUSH_NEIGHBOR_COUNT,
MONITOR_FLUSH_NEIGHBOR_PAGES, n);
}
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
return count;
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
}
/*******************************************************************//**
This utility moves the uncompressed frames of pages to the free list.
Note that this function does not actually flush any data to disk. It
just detaches the uncompressed frames from the compressed pages at the
tail of the unzip_LRU and puts those freed frames in the free list.
Note that it is a best effort attempt and it is not guaranteed that
after a call to this function there will be 'max' blocks in the free
list.
MDEV-15058: Deprecate and ignore innodb_buffer_pool_instances Our benchmarking efforts indicate that the reasons for splitting the buf_pool in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 have mostly gone away, possibly as a result of mysql/mysql-server@ce6109ebfdedfdf185e391a0c97dc6d33867ed78 or similar work. Only in one write-heavy benchmark where the working set size is ten times the buffer pool size, the buf_pool->mutex would be less contended with 4 buffer pool instances than with 1 instance, in buf_page_io_complete(). That contention could be alleviated further by making more use of std::atomic and by splitting buf_pool_t::mutex further (MDEV-15053). We will deprecate and ignore the following parameters: innodb_buffer_pool_instances innodb_page_cleaners There will be only one buffer pool and one page cleaner task. In a number of INFORMATION_SCHEMA views, columns that indicated the buffer pool instance will be removed: information_schema.innodb_buffer_page.pool_id information_schema.innodb_buffer_page_lru.pool_id information_schema.innodb_buffer_pool_stats.pool_id information_schema.innodb_cmpmem.buffer_pool_instance information_schema.innodb_cmpmem_reset.buffer_pool_instance
2020-02-12 13:45:21 +01:00
@param[in] max desired number of blocks in the free_list
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
@return number of blocks moved to the free list. */
MDEV-15058: Deprecate and ignore innodb_buffer_pool_instances Our benchmarking efforts indicate that the reasons for splitting the buf_pool in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 have mostly gone away, possibly as a result of mysql/mysql-server@ce6109ebfdedfdf185e391a0c97dc6d33867ed78 or similar work. Only in one write-heavy benchmark where the working set size is ten times the buffer pool size, the buf_pool->mutex would be less contended with 4 buffer pool instances than with 1 instance, in buf_page_io_complete(). That contention could be alleviated further by making more use of std::atomic and by splitting buf_pool_t::mutex further (MDEV-15053). We will deprecate and ignore the following parameters: innodb_buffer_pool_instances innodb_page_cleaners There will be only one buffer pool and one page cleaner task. In a number of INFORMATION_SCHEMA views, columns that indicated the buffer pool instance will be removed: information_schema.innodb_buffer_page.pool_id information_schema.innodb_buffer_page_lru.pool_id information_schema.innodb_buffer_pool_stats.pool_id information_schema.innodb_cmpmem.buffer_pool_instance information_schema.innodb_cmpmem_reset.buffer_pool_instance
2020-02-12 13:45:21 +01:00
static ulint buf_free_from_unzip_LRU_list_batch(ulint max)
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
{
ulint scanned = 0;
ulint count = 0;
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
mysql_mutex_assert_owner(&buf_pool.mutex);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-21962 Allocate buf_pool statically Thanks to MDEV-15058, there is only one InnoDB buffer pool. Allocating buf_pool statically removes one level of pointer indirection and makes code more readable, and removes the awkward initialization of some buf_pool members. While doing this, we will also declare some buf_pool_t data members private and replace some functions with member functions. This is mostly affecting buffer pool resizing. This is not aiming to be a complete rewrite of buf_pool_t to a proper class. Most of the buffer pool interface, such as buf_page_get_gen(), will remain in the C programming style for now. buf_pool_t::withdrawing: Replaces buf_pool_withdrawing. buf_pool_t::withdraw_clock_: Replaces buf_withdraw_clock. buf_pool_t::create(): Repalces buf_pool_init(). buf_pool_t::close(): Replaces buf_pool_free(). buf_bool_t::will_be_withdrawn(): Replaces buf_block_will_be_withdrawn(), buf_frame_will_be_withdrawn(). buf_pool_t::clear_hash_index(): Replaces buf_pool_clear_hash_index(). buf_pool_t::get_n_pages(): Replaces buf_pool_get_n_pages(). buf_pool_t::validate(): Replaces buf_validate(). buf_pool_t::print(): Replaces buf_print(). buf_pool_t::block_from_ahi(): Replaces buf_block_from_ahi(). buf_pool_t::is_block_field(): Replaces buf_pointer_is_block_field(). buf_pool_t::is_block_mutex(): Replaces buf_pool_is_block_mutex(). buf_pool_t::is_block_lock(): Replaces buf_pool_is_block_lock(). buf_pool_t::is_obsolete(): Replaces buf_pool_is_obsolete(). buf_pool_t::io_buf: Make default-constructible. buf_pool_t::io_buf::create(): Delayed 'constructor' buf_pool_t::io_buf::close(): Early 'destructor' HazardPointer: Make default-constructible. Define all member functions inline, also for derived classes.
2020-03-18 20:48:00 +01:00
buf_block_t* block = UT_LIST_GET_LAST(buf_pool.unzip_LRU);
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
while (block
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
&& count < max
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
&& UT_LIST_GET_LEN(buf_pool.free) < srv_LRU_scan_depth
&& UT_LIST_GET_LEN(buf_pool.unzip_LRU)
> UT_LIST_GET_LEN(buf_pool.LRU) / 10) {
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
++scanned;
if (buf_LRU_free_page(&block->page, false)) {
MDEV-21962 Allocate buf_pool statically Thanks to MDEV-15058, there is only one InnoDB buffer pool. Allocating buf_pool statically removes one level of pointer indirection and makes code more readable, and removes the awkward initialization of some buf_pool members. While doing this, we will also declare some buf_pool_t data members private and replace some functions with member functions. This is mostly affecting buffer pool resizing. This is not aiming to be a complete rewrite of buf_pool_t to a proper class. Most of the buffer pool interface, such as buf_page_get_gen(), will remain in the C programming style for now. buf_pool_t::withdrawing: Replaces buf_pool_withdrawing. buf_pool_t::withdraw_clock_: Replaces buf_withdraw_clock. buf_pool_t::create(): Repalces buf_pool_init(). buf_pool_t::close(): Replaces buf_pool_free(). buf_bool_t::will_be_withdrawn(): Replaces buf_block_will_be_withdrawn(), buf_frame_will_be_withdrawn(). buf_pool_t::clear_hash_index(): Replaces buf_pool_clear_hash_index(). buf_pool_t::get_n_pages(): Replaces buf_pool_get_n_pages(). buf_pool_t::validate(): Replaces buf_validate(). buf_pool_t::print(): Replaces buf_print(). buf_pool_t::block_from_ahi(): Replaces buf_block_from_ahi(). buf_pool_t::is_block_field(): Replaces buf_pointer_is_block_field(). buf_pool_t::is_block_mutex(): Replaces buf_pool_is_block_mutex(). buf_pool_t::is_block_lock(): Replaces buf_pool_is_block_lock(). buf_pool_t::is_obsolete(): Replaces buf_pool_is_obsolete(). buf_pool_t::io_buf: Make default-constructible. buf_pool_t::io_buf::create(): Delayed 'constructor' buf_pool_t::io_buf::close(): Early 'destructor' HazardPointer: Make default-constructible. Define all member functions inline, also for derived classes.
2020-03-18 20:48:00 +01:00
/* Block was freed. buf_pool.mutex potentially
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
released and reacquired */
++count;
MDEV-21962 Allocate buf_pool statically Thanks to MDEV-15058, there is only one InnoDB buffer pool. Allocating buf_pool statically removes one level of pointer indirection and makes code more readable, and removes the awkward initialization of some buf_pool members. While doing this, we will also declare some buf_pool_t data members private and replace some functions with member functions. This is mostly affecting buffer pool resizing. This is not aiming to be a complete rewrite of buf_pool_t to a proper class. Most of the buffer pool interface, such as buf_page_get_gen(), will remain in the C programming style for now. buf_pool_t::withdrawing: Replaces buf_pool_withdrawing. buf_pool_t::withdraw_clock_: Replaces buf_withdraw_clock. buf_pool_t::create(): Repalces buf_pool_init(). buf_pool_t::close(): Replaces buf_pool_free(). buf_bool_t::will_be_withdrawn(): Replaces buf_block_will_be_withdrawn(), buf_frame_will_be_withdrawn(). buf_pool_t::clear_hash_index(): Replaces buf_pool_clear_hash_index(). buf_pool_t::get_n_pages(): Replaces buf_pool_get_n_pages(). buf_pool_t::validate(): Replaces buf_validate(). buf_pool_t::print(): Replaces buf_print(). buf_pool_t::block_from_ahi(): Replaces buf_block_from_ahi(). buf_pool_t::is_block_field(): Replaces buf_pointer_is_block_field(). buf_pool_t::is_block_mutex(): Replaces buf_pool_is_block_mutex(). buf_pool_t::is_block_lock(): Replaces buf_pool_is_block_lock(). buf_pool_t::is_obsolete(): Replaces buf_pool_is_obsolete(). buf_pool_t::io_buf: Make default-constructible. buf_pool_t::io_buf::create(): Delayed 'constructor' buf_pool_t::io_buf::close(): Early 'destructor' HazardPointer: Make default-constructible. Define all member functions inline, also for derived classes.
2020-03-18 20:48:00 +01:00
block = UT_LIST_GET_LAST(buf_pool.unzip_LRU);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
} else {
block = UT_LIST_GET_PREV(unzip_LRU, block);
}
}
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
mysql_mutex_assert_owner(&buf_pool.mutex);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
if (scanned) {
MONITOR_INC_VALUE_CUMULATIVE(
MONITOR_LRU_BATCH_SCANNED,
MONITOR_LRU_BATCH_SCANNED_NUM_CALL,
MONITOR_LRU_BATCH_SCANNED_PER_CALL,
scanned);
}
return(count);
}
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
/** Start writing out pages for a tablespace.
@param id tablespace identifier
MDEV-27985 buf_flush_freed_pages() causes InnoDB to hang buf_flush_freed_pages(): Assert that neither buf_pool.mutex nor buf_pool.flush_list_mutex are held. Simplify the loops. Return the tablespace and the number of pages written or punched. buf_flush_LRU_list_batch(), buf_do_flush_list_batch(): Release buf_pool.mutex before invoking buf_flush_space(). buf_flush_list_space(): Acquire the mutexes only after invoking buf_flush_freed_pages(). Reviewed by: Thirunarayanan Balathandayuthapani
2022-03-15 13:44:22 +01:00
@return tablespace and number of pages written */
static std::pair<fil_space_t*, uint32_t> buf_flush_space(const uint32_t id)
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
{
MDEV-27985 buf_flush_freed_pages() causes InnoDB to hang buf_flush_freed_pages(): Assert that neither buf_pool.mutex nor buf_pool.flush_list_mutex are held. Simplify the loops. Return the tablespace and the number of pages written or punched. buf_flush_LRU_list_batch(), buf_do_flush_list_batch(): Release buf_pool.mutex before invoking buf_flush_space(). buf_flush_list_space(): Acquire the mutexes only after invoking buf_flush_freed_pages(). Reviewed by: Thirunarayanan Balathandayuthapani
2022-03-15 13:44:22 +01:00
if (fil_space_t *space= fil_space_t::get(id))
Cleanup: Make fil_space_t::freed_ranges private fil_space_t::is_freed(): Check if a page is in freed_ranges. fil_space_t::flush_freed(): Replaces buf_flush_freed_pages().
2022-06-06 10:55:29 +02:00
return {space, space->flush_freed(true)};
MDEV-27985 buf_flush_freed_pages() causes InnoDB to hang buf_flush_freed_pages(): Assert that neither buf_pool.mutex nor buf_pool.flush_list_mutex are held. Simplify the loops. Return the tablespace and the number of pages written or punched. buf_flush_LRU_list_batch(), buf_do_flush_list_batch(): Release buf_pool.mutex before invoking buf_flush_space(). buf_flush_list_space(): Acquire the mutexes only after invoking buf_flush_freed_pages(). Reviewed by: Thirunarayanan Balathandayuthapani
2022-03-15 13:44:22 +01:00
return {nullptr, 0};
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
}
struct flush_counters_t
{
/** number of dirty pages flushed */
ulint flushed;
/** number of clean pages evicted */
ulint evicted;
};
Merge InnoDB 5.7 from mysql-5.7.14. Contains also: MDEV-10549 mysqld: sql/handler.cc:2692: int handler::ha_index_first(uchar*): Assertion `table_share->tmp_table != NO_TMP_TABLE || m_lock_type != 2' failed. (branch bb-10.2-jan) Unlike MySQL, InnoDB still uses THR_LOCK in MariaDB MDEV-10548 Some of the debug sync waits do not work with InnoDB 5.7 (branch bb-10.2-jan) enable tests that were fixed in MDEV-10549 MDEV-10548 Some of the debug sync waits do not work with InnoDB 5.7 (branch bb-10.2-jan) fix main.innodb_mysql_sync - re-enable online alter for partitioned innodb tables
2016-09-06 08:43:16 +02:00
MDEV-23855: Remove fil_system.LRU and reduce fil_system.mutex contention Also fixes MDEV-23929: innodb_flush_neighbors is not being ignored for system tablespace on SSD When the maximum configured number of file is exceeded, InnoDB will close data files. We used to maintain a fil_system.LRU list and a counter fil_node_t::n_pending to achieve this, at the huge cost of multiple fil_system.mutex operations per I/O operation. fil_node_open_file_low(): Implement a FIFO replacement policy: The last opened file will be moved to the end of fil_system.space_list, and files will be closed from the start of the list. However, we will not move tablespaces in fil_system.space_list while i_s_tablespaces_encryption_fill_table() is executing (producing output for INFORMATION_SCHEMA.INNODB_TABLESPACES_ENCRYPTION) because it may cause information of some tablespaces to go missing. We also avoid this in mariabackup --backup because datafiles_iter_next() assumes that the ordering is not changed. IORequest: Fold more parameters to IORequest::type. fil_space_t::io(): Replaces fil_io(). fil_space_t::flush(): Replaces fil_flush(). OS_AIO_IBUF: Remove. We will always issue synchronous reads of the change buffer pages in buf_read_page_low(). We will always ignore some errors for background reads. This should reduce fil_system.mutex contention a little. fil_node_t::complete_write(): Replaces fil_node_t::complete_io(). On both read and write completion, fil_space_t::release_for_io() will have to be called. fil_space_t::io(): Do not acquire fil_system.mutex in the normal code path. xb_delta_open_matching_space(): Do not try to open the system tablespace which was already opened. This fixes a file sharing violation in mariabackup --prepare --incremental. Reviewed by: Vladislav Vaintroub
2020-10-26 14:59:30 +01:00
/** Try to discard a dirty page.
@param bpage dirty page whose tablespace is not accessible */
static void buf_flush_discard_page(buf_page_t *bpage)
{
mysql_mutex_assert_owner(&buf_pool.mutex);
mysql_mutex_assert_not_owner(&buf_pool.flush_list_mutex);
ut_ad(bpage->in_file());
ut_ad(bpage->oldest_modification());
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
if (!bpage->lock.u_lock_try(false))
return;
MDEV-23855: Remove fil_system.LRU and reduce fil_system.mutex contention Also fixes MDEV-23929: innodb_flush_neighbors is not being ignored for system tablespace on SSD When the maximum configured number of file is exceeded, InnoDB will close data files. We used to maintain a fil_system.LRU list and a counter fil_node_t::n_pending to achieve this, at the huge cost of multiple fil_system.mutex operations per I/O operation. fil_node_open_file_low(): Implement a FIFO replacement policy: The last opened file will be moved to the end of fil_system.space_list, and files will be closed from the start of the list. However, we will not move tablespaces in fil_system.space_list while i_s_tablespaces_encryption_fill_table() is executing (producing output for INFORMATION_SCHEMA.INNODB_TABLESPACES_ENCRYPTION) because it may cause information of some tablespaces to go missing. We also avoid this in mariabackup --backup because datafiles_iter_next() assumes that the ordering is not changed. IORequest: Fold more parameters to IORequest::type. fil_space_t::io(): Replaces fil_io(). fil_space_t::flush(): Replaces fil_flush(). OS_AIO_IBUF: Remove. We will always issue synchronous reads of the change buffer pages in buf_read_page_low(). We will always ignore some errors for background reads. This should reduce fil_system.mutex contention a little. fil_node_t::complete_write(): Replaces fil_node_t::complete_io(). On both read and write completion, fil_space_t::release_for_io() will have to be called. fil_space_t::io(): Do not acquire fil_system.mutex in the normal code path. xb_delta_open_matching_space(): Do not try to open the system tablespace which was already opened. This fixes a file sharing violation in mariabackup --prepare --incremental. Reviewed by: Vladislav Vaintroub
2020-10-26 14:59:30 +01:00
mysql_mutex_lock(&buf_pool.flush_list_mutex);
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
buf_pool.delete_from_flush_list(bpage);
MDEV-23855: Remove fil_system.LRU and reduce fil_system.mutex contention Also fixes MDEV-23929: innodb_flush_neighbors is not being ignored for system tablespace on SSD When the maximum configured number of file is exceeded, InnoDB will close data files. We used to maintain a fil_system.LRU list and a counter fil_node_t::n_pending to achieve this, at the huge cost of multiple fil_system.mutex operations per I/O operation. fil_node_open_file_low(): Implement a FIFO replacement policy: The last opened file will be moved to the end of fil_system.space_list, and files will be closed from the start of the list. However, we will not move tablespaces in fil_system.space_list while i_s_tablespaces_encryption_fill_table() is executing (producing output for INFORMATION_SCHEMA.INNODB_TABLESPACES_ENCRYPTION) because it may cause information of some tablespaces to go missing. We also avoid this in mariabackup --backup because datafiles_iter_next() assumes that the ordering is not changed. IORequest: Fold more parameters to IORequest::type. fil_space_t::io(): Replaces fil_io(). fil_space_t::flush(): Replaces fil_flush(). OS_AIO_IBUF: Remove. We will always issue synchronous reads of the change buffer pages in buf_read_page_low(). We will always ignore some errors for background reads. This should reduce fil_system.mutex contention a little. fil_node_t::complete_write(): Replaces fil_node_t::complete_io(). On both read and write completion, fil_space_t::release_for_io() will have to be called. fil_space_t::io(): Do not acquire fil_system.mutex in the normal code path. xb_delta_open_matching_space(): Do not try to open the system tablespace which was already opened. This fixes a file sharing violation in mariabackup --prepare --incremental. Reviewed by: Vladislav Vaintroub
2020-10-26 14:59:30 +01:00
mysql_mutex_unlock(&buf_pool.flush_list_mutex);
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
ut_d(const auto state= bpage->state());
ut_ad(state == buf_page_t::FREED || state == buf_page_t::UNFIXED ||
state == buf_page_t::IBUF_EXIST || state == buf_page_t::REINIT);
bpage->lock.u_unlock();
MDEV-23855: Remove fil_system.LRU and reduce fil_system.mutex contention Also fixes MDEV-23929: innodb_flush_neighbors is not being ignored for system tablespace on SSD When the maximum configured number of file is exceeded, InnoDB will close data files. We used to maintain a fil_system.LRU list and a counter fil_node_t::n_pending to achieve this, at the huge cost of multiple fil_system.mutex operations per I/O operation. fil_node_open_file_low(): Implement a FIFO replacement policy: The last opened file will be moved to the end of fil_system.space_list, and files will be closed from the start of the list. However, we will not move tablespaces in fil_system.space_list while i_s_tablespaces_encryption_fill_table() is executing (producing output for INFORMATION_SCHEMA.INNODB_TABLESPACES_ENCRYPTION) because it may cause information of some tablespaces to go missing. We also avoid this in mariabackup --backup because datafiles_iter_next() assumes that the ordering is not changed. IORequest: Fold more parameters to IORequest::type. fil_space_t::io(): Replaces fil_io(). fil_space_t::flush(): Replaces fil_flush(). OS_AIO_IBUF: Remove. We will always issue synchronous reads of the change buffer pages in buf_read_page_low(). We will always ignore some errors for background reads. This should reduce fil_system.mutex contention a little. fil_node_t::complete_write(): Replaces fil_node_t::complete_io(). On both read and write completion, fil_space_t::release_for_io() will have to be called. fil_space_t::io(): Do not acquire fil_system.mutex in the normal code path. xb_delta_open_matching_space(): Do not try to open the system tablespace which was already opened. This fixes a file sharing violation in mariabackup --prepare --incremental. Reviewed by: Vladislav Vaintroub
2020-10-26 14:59:30 +01:00
buf_LRU_free_page(bpage, true);
}
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
/** Flush dirty blocks from the end of the LRU list.
@param max maximum number of blocks to make available in buf_pool.free
@param n counts of flushed and evicted pages */
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
static void buf_flush_LRU_list_batch(ulint max, flush_counters_t *n)
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
{
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
ulint scanned= 0;
ulint free_limit= srv_LRU_scan_depth;
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
mysql_mutex_assert_owner(&buf_pool.mutex);
MDEV-27891: SIGSEGV in InnoDB buffer pool resize During an increase in resize, the new curr_size got a value less than old_size. As n_chunks_new and n_chunks have a strong correlation to the resizing operation in progress, we can use them and remove the need for old_size. For convienece the n_chunks_new < n_chunks is now the is_shrinking function. The volatile compiler optimization on n_chunks{,_new} is removed as real mutex uses are needed. Other n_chunks_new/n_chunks methods: n_chunks_new and n_chunks almost always read and altered under the pool mutex. Exceptions are: * i_s_innodb_buffer_page_fill, * buf_pool_t::is_uncompressed (via is_blocked_field) These need reexamining for the need of a mutex, however comments indicates this already. get_n_pages has uses in buffer pool load, recover log memory exhaustion estimates and innodb status so take the minimum number of chunks for safety. The buf_pool_t::running_out function also uses curr_size/old_size. We replace this hot function calculation with just n_chunks_new. This is the new size of the chunks before the resizing occurs. If we are resizing down, we've already got the case we had previously (as the minimum). If we are resizing upwards, we are taking an optimistic view that there will be buffer chunks available for locks. As this memory allocation is occurring immediately next the resizing function it seems likely. Compiler hint UNIV_UNLIKELY removed to leave it to the branch predictor to make an informed decision. Added test case of a smaller size than the Marko/Roel original in JIRA reducing the size to 256M. SEGV hits roughly 1/10 times but its better than a 21G memory size. Reviewer: Marko
2022-02-22 07:42:59 +01:00
if (buf_pool.withdraw_target && buf_pool.is_shrinking())
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
free_limit+= buf_pool.withdraw_target - UT_LIST_GET_LEN(buf_pool.withdraw);
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
const auto neighbors= UT_LIST_GET_LEN(buf_pool.LRU) < BUF_LRU_OLD_MIN_LEN
? 0 : srv_flush_neighbors;
fil_space_t *space= nullptr;
MDEV-23855: Remove fil_system.LRU and reduce fil_system.mutex contention Also fixes MDEV-23929: innodb_flush_neighbors is not being ignored for system tablespace on SSD When the maximum configured number of file is exceeded, InnoDB will close data files. We used to maintain a fil_system.LRU list and a counter fil_node_t::n_pending to achieve this, at the huge cost of multiple fil_system.mutex operations per I/O operation. fil_node_open_file_low(): Implement a FIFO replacement policy: The last opened file will be moved to the end of fil_system.space_list, and files will be closed from the start of the list. However, we will not move tablespaces in fil_system.space_list while i_s_tablespaces_encryption_fill_table() is executing (producing output for INFORMATION_SCHEMA.INNODB_TABLESPACES_ENCRYPTION) because it may cause information of some tablespaces to go missing. We also avoid this in mariabackup --backup because datafiles_iter_next() assumes that the ordering is not changed. IORequest: Fold more parameters to IORequest::type. fil_space_t::io(): Replaces fil_io(). fil_space_t::flush(): Replaces fil_flush(). OS_AIO_IBUF: Remove. We will always issue synchronous reads of the change buffer pages in buf_read_page_low(). We will always ignore some errors for background reads. This should reduce fil_system.mutex contention a little. fil_node_t::complete_write(): Replaces fil_node_t::complete_io(). On both read and write completion, fil_space_t::release_for_io() will have to be called. fil_space_t::io(): Do not acquire fil_system.mutex in the normal code path. xb_delta_open_matching_space(): Do not try to open the system tablespace which was already opened. This fixes a file sharing violation in mariabackup --prepare --incremental. Reviewed by: Vladislav Vaintroub
2020-10-26 14:59:30 +01:00
uint32_t last_space_id= FIL_NULL;
static_assert(FIL_NULL > SRV_TMP_SPACE_ID, "consistency");
static_assert(FIL_NULL > SRV_SPACE_ID_UPPER_BOUND, "consistency");
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
for (buf_page_t *bpage= UT_LIST_GET_LAST(buf_pool.LRU);
MDEV-26784 [Warning] InnoDB: Difficult to find free blocks in the buffer pool Problem: ======= InnoDB ran out of memory during recovery and it fails to flush the dirty LRU blocks. The reason is that buffer pool can ran out before the LRU list length reaches BUF_LRU_OLD_MIN_LEN(256) threshold. Fix: ==== During recovery, InnoDB should write out and evict all dirty blocks.
2022-01-20 12:00:18 +01:00
bpage &&
((UT_LIST_GET_LEN(buf_pool.LRU) > BUF_LRU_MIN_LEN &&
UT_LIST_GET_LEN(buf_pool.free) < free_limit &&
n->flushed + n->evicted < max) ||
recv_recovery_is_on()); ++scanned)
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
{
buf_page_t *prev= UT_LIST_GET_PREV(LRU, bpage);
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
const lsn_t oldest_modification= bpage->oldest_modification();
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
buf_pool.lru_hp.set(prev);
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
const auto state= bpage->state();
ut_ad(state >= buf_page_t::FREED);
ut_ad(bpage->in_LRU_list);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
if (oldest_modification <= 1)
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
{
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
if (state != buf_page_t::FREED &&
(state >= buf_page_t::READ_FIX || (~buf_page_t::LRU_MASK & state)))
goto must_skip;
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
if (buf_LRU_free_page(bpage, true))
++n->evicted;
}
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
else if (state < buf_page_t::READ_FIX)
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
{
/* Block is ready for flush. Dispatch an IO request. The IO
helper thread will put it on free list in IO completion routine. */
const page_id_t page_id(bpage->id());
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
const uint32_t space_id= page_id.space();
if (!space || space->id != space_id)
{
MDEV-23855: Remove fil_system.LRU and reduce fil_system.mutex contention Also fixes MDEV-23929: innodb_flush_neighbors is not being ignored for system tablespace on SSD When the maximum configured number of file is exceeded, InnoDB will close data files. We used to maintain a fil_system.LRU list and a counter fil_node_t::n_pending to achieve this, at the huge cost of multiple fil_system.mutex operations per I/O operation. fil_node_open_file_low(): Implement a FIFO replacement policy: The last opened file will be moved to the end of fil_system.space_list, and files will be closed from the start of the list. However, we will not move tablespaces in fil_system.space_list while i_s_tablespaces_encryption_fill_table() is executing (producing output for INFORMATION_SCHEMA.INNODB_TABLESPACES_ENCRYPTION) because it may cause information of some tablespaces to go missing. We also avoid this in mariabackup --backup because datafiles_iter_next() assumes that the ordering is not changed. IORequest: Fold more parameters to IORequest::type. fil_space_t::io(): Replaces fil_io(). fil_space_t::flush(): Replaces fil_flush(). OS_AIO_IBUF: Remove. We will always issue synchronous reads of the change buffer pages in buf_read_page_low(). We will always ignore some errors for background reads. This should reduce fil_system.mutex contention a little. fil_node_t::complete_write(): Replaces fil_node_t::complete_io(). On both read and write completion, fil_space_t::release_for_io() will have to be called. fil_space_t::io(): Do not acquire fil_system.mutex in the normal code path. xb_delta_open_matching_space(): Do not try to open the system tablespace which was already opened. This fixes a file sharing violation in mariabackup --prepare --incremental. Reviewed by: Vladislav Vaintroub
2020-10-26 14:59:30 +01:00
if (last_space_id != space_id)
{
MDEV-27985 buf_flush_freed_pages() causes InnoDB to hang buf_flush_freed_pages(): Assert that neither buf_pool.mutex nor buf_pool.flush_list_mutex are held. Simplify the loops. Return the tablespace and the number of pages written or punched. buf_flush_LRU_list_batch(), buf_do_flush_list_batch(): Release buf_pool.mutex before invoking buf_flush_space(). buf_flush_list_space(): Acquire the mutexes only after invoking buf_flush_freed_pages(). Reviewed by: Thirunarayanan Balathandayuthapani
2022-03-15 13:44:22 +01:00
buf_pool.lru_hp.set(bpage);
mysql_mutex_unlock(&buf_pool.mutex);
MDEV-23855: Remove fil_system.LRU and reduce fil_system.mutex contention Also fixes MDEV-23929: innodb_flush_neighbors is not being ignored for system tablespace on SSD When the maximum configured number of file is exceeded, InnoDB will close data files. We used to maintain a fil_system.LRU list and a counter fil_node_t::n_pending to achieve this, at the huge cost of multiple fil_system.mutex operations per I/O operation. fil_node_open_file_low(): Implement a FIFO replacement policy: The last opened file will be moved to the end of fil_system.space_list, and files will be closed from the start of the list. However, we will not move tablespaces in fil_system.space_list while i_s_tablespaces_encryption_fill_table() is executing (producing output for INFORMATION_SCHEMA.INNODB_TABLESPACES_ENCRYPTION) because it may cause information of some tablespaces to go missing. We also avoid this in mariabackup --backup because datafiles_iter_next() assumes that the ordering is not changed. IORequest: Fold more parameters to IORequest::type. fil_space_t::io(): Replaces fil_io(). fil_space_t::flush(): Replaces fil_flush(). OS_AIO_IBUF: Remove. We will always issue synchronous reads of the change buffer pages in buf_read_page_low(). We will always ignore some errors for background reads. This should reduce fil_system.mutex contention a little. fil_node_t::complete_write(): Replaces fil_node_t::complete_io(). On both read and write completion, fil_space_t::release_for_io() will have to be called. fil_space_t::io(): Do not acquire fil_system.mutex in the normal code path. xb_delta_open_matching_space(): Do not try to open the system tablespace which was already opened. This fixes a file sharing violation in mariabackup --prepare --incremental. Reviewed by: Vladislav Vaintroub
2020-10-26 14:59:30 +01:00
if (space)
MDEV-23855: Shrink fil_space_t Merge n_pending_ios, n_pending_ops to std::atomic<uint32_t> n_pending. Change some more fil_space_t members to uint32_t to reduce the memory footprint. fil_space_t::add(), fil_ibd_create(): Attach the already opened handle to the tablespace, and enforce the fil_system.n_open limit. dict_boot(): Initialize fil_system.max_assigned_id. srv_boot(): Call srv_thread_pool_init() before anything else, so that files should be opened in the correct mode on Windows. fil_ibd_create(): Create the file in OS_FILE_AIO mode, just like fil_node_open_file_low() does it. dict_table_t::is_accessible(): Replaces fil_table_accessible(). Reviewed by: Vladislav Vaintroub
2020-10-26 15:04:12 +01:00
space->release();
MDEV-27985 buf_flush_freed_pages() causes InnoDB to hang buf_flush_freed_pages(): Assert that neither buf_pool.mutex nor buf_pool.flush_list_mutex are held. Simplify the loops. Return the tablespace and the number of pages written or punched. buf_flush_LRU_list_batch(), buf_do_flush_list_batch(): Release buf_pool.mutex before invoking buf_flush_space(). buf_flush_list_space(): Acquire the mutexes only after invoking buf_flush_freed_pages(). Reviewed by: Thirunarayanan Balathandayuthapani
2022-03-15 13:44:22 +01:00
auto p= buf_flush_space(space_id);
space= p.first;
MDEV-23855: Remove fil_system.LRU and reduce fil_system.mutex contention Also fixes MDEV-23929: innodb_flush_neighbors is not being ignored for system tablespace on SSD When the maximum configured number of file is exceeded, InnoDB will close data files. We used to maintain a fil_system.LRU list and a counter fil_node_t::n_pending to achieve this, at the huge cost of multiple fil_system.mutex operations per I/O operation. fil_node_open_file_low(): Implement a FIFO replacement policy: The last opened file will be moved to the end of fil_system.space_list, and files will be closed from the start of the list. However, we will not move tablespaces in fil_system.space_list while i_s_tablespaces_encryption_fill_table() is executing (producing output for INFORMATION_SCHEMA.INNODB_TABLESPACES_ENCRYPTION) because it may cause information of some tablespaces to go missing. We also avoid this in mariabackup --backup because datafiles_iter_next() assumes that the ordering is not changed. IORequest: Fold more parameters to IORequest::type. fil_space_t::io(): Replaces fil_io(). fil_space_t::flush(): Replaces fil_flush(). OS_AIO_IBUF: Remove. We will always issue synchronous reads of the change buffer pages in buf_read_page_low(). We will always ignore some errors for background reads. This should reduce fil_system.mutex contention a little. fil_node_t::complete_write(): Replaces fil_node_t::complete_io(). On both read and write completion, fil_space_t::release_for_io() will have to be called. fil_space_t::io(): Do not acquire fil_system.mutex in the normal code path. xb_delta_open_matching_space(): Do not try to open the system tablespace which was already opened. This fixes a file sharing violation in mariabackup --prepare --incremental. Reviewed by: Vladislav Vaintroub
2020-10-26 14:59:30 +01:00
last_space_id= space_id;
MDEV-27985 buf_flush_freed_pages() causes InnoDB to hang buf_flush_freed_pages(): Assert that neither buf_pool.mutex nor buf_pool.flush_list_mutex are held. Simplify the loops. Return the tablespace and the number of pages written or punched. buf_flush_LRU_list_batch(), buf_do_flush_list_batch(): Release buf_pool.mutex before invoking buf_flush_space(). buf_flush_list_space(): Acquire the mutexes only after invoking buf_flush_freed_pages(). Reviewed by: Thirunarayanan Balathandayuthapani
2022-03-15 13:44:22 +01:00
mysql_mutex_lock(&buf_pool.mutex);
if (p.second)
buf_pool.stat.n_pages_written+= p.second;
goto retry;
MDEV-23855: Remove fil_system.LRU and reduce fil_system.mutex contention Also fixes MDEV-23929: innodb_flush_neighbors is not being ignored for system tablespace on SSD When the maximum configured number of file is exceeded, InnoDB will close data files. We used to maintain a fil_system.LRU list and a counter fil_node_t::n_pending to achieve this, at the huge cost of multiple fil_system.mutex operations per I/O operation. fil_node_open_file_low(): Implement a FIFO replacement policy: The last opened file will be moved to the end of fil_system.space_list, and files will be closed from the start of the list. However, we will not move tablespaces in fil_system.space_list while i_s_tablespaces_encryption_fill_table() is executing (producing output for INFORMATION_SCHEMA.INNODB_TABLESPACES_ENCRYPTION) because it may cause information of some tablespaces to go missing. We also avoid this in mariabackup --backup because datafiles_iter_next() assumes that the ordering is not changed. IORequest: Fold more parameters to IORequest::type. fil_space_t::io(): Replaces fil_io(). fil_space_t::flush(): Replaces fil_flush(). OS_AIO_IBUF: Remove. We will always issue synchronous reads of the change buffer pages in buf_read_page_low(). We will always ignore some errors for background reads. This should reduce fil_system.mutex contention a little. fil_node_t::complete_write(): Replaces fil_node_t::complete_io(). On both read and write completion, fil_space_t::release_for_io() will have to be called. fil_space_t::io(): Do not acquire fil_system.mutex in the normal code path. xb_delta_open_matching_space(): Do not try to open the system tablespace which was already opened. This fixes a file sharing violation in mariabackup --prepare --incremental. Reviewed by: Vladislav Vaintroub
2020-10-26 14:59:30 +01:00
}
else
ut_ad(!space);
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
}
MDEV-23855: Remove fil_system.LRU and reduce fil_system.mutex contention Also fixes MDEV-23929: innodb_flush_neighbors is not being ignored for system tablespace on SSD When the maximum configured number of file is exceeded, InnoDB will close data files. We used to maintain a fil_system.LRU list and a counter fil_node_t::n_pending to achieve this, at the huge cost of multiple fil_system.mutex operations per I/O operation. fil_node_open_file_low(): Implement a FIFO replacement policy: The last opened file will be moved to the end of fil_system.space_list, and files will be closed from the start of the list. However, we will not move tablespaces in fil_system.space_list while i_s_tablespaces_encryption_fill_table() is executing (producing output for INFORMATION_SCHEMA.INNODB_TABLESPACES_ENCRYPTION) because it may cause information of some tablespaces to go missing. We also avoid this in mariabackup --backup because datafiles_iter_next() assumes that the ordering is not changed. IORequest: Fold more parameters to IORequest::type. fil_space_t::io(): Replaces fil_io(). fil_space_t::flush(): Replaces fil_flush(). OS_AIO_IBUF: Remove. We will always issue synchronous reads of the change buffer pages in buf_read_page_low(). We will always ignore some errors for background reads. This should reduce fil_system.mutex contention a little. fil_node_t::complete_write(): Replaces fil_node_t::complete_io(). On both read and write completion, fil_space_t::release_for_io() will have to be called. fil_space_t::io(): Do not acquire fil_system.mutex in the normal code path. xb_delta_open_matching_space(): Do not try to open the system tablespace which was already opened. This fixes a file sharing violation in mariabackup --prepare --incremental. Reviewed by: Vladislav Vaintroub
2020-10-26 14:59:30 +01:00
else if (space->is_stopping())
{
MDEV-23855: Shrink fil_space_t Merge n_pending_ios, n_pending_ops to std::atomic<uint32_t> n_pending. Change some more fil_space_t members to uint32_t to reduce the memory footprint. fil_space_t::add(), fil_ibd_create(): Attach the already opened handle to the tablespace, and enforce the fil_system.n_open limit. dict_boot(): Initialize fil_system.max_assigned_id. srv_boot(): Call srv_thread_pool_init() before anything else, so that files should be opened in the correct mode on Windows. fil_ibd_create(): Create the file in OS_FILE_AIO mode, just like fil_node_open_file_low() does it. dict_table_t::is_accessible(): Replaces fil_table_accessible(). Reviewed by: Vladislav Vaintroub
2020-10-26 15:04:12 +01:00
space->release();
MDEV-23855: Remove fil_system.LRU and reduce fil_system.mutex contention Also fixes MDEV-23929: innodb_flush_neighbors is not being ignored for system tablespace on SSD When the maximum configured number of file is exceeded, InnoDB will close data files. We used to maintain a fil_system.LRU list and a counter fil_node_t::n_pending to achieve this, at the huge cost of multiple fil_system.mutex operations per I/O operation. fil_node_open_file_low(): Implement a FIFO replacement policy: The last opened file will be moved to the end of fil_system.space_list, and files will be closed from the start of the list. However, we will not move tablespaces in fil_system.space_list while i_s_tablespaces_encryption_fill_table() is executing (producing output for INFORMATION_SCHEMA.INNODB_TABLESPACES_ENCRYPTION) because it may cause information of some tablespaces to go missing. We also avoid this in mariabackup --backup because datafiles_iter_next() assumes that the ordering is not changed. IORequest: Fold more parameters to IORequest::type. fil_space_t::io(): Replaces fil_io(). fil_space_t::flush(): Replaces fil_flush(). OS_AIO_IBUF: Remove. We will always issue synchronous reads of the change buffer pages in buf_read_page_low(). We will always ignore some errors for background reads. This should reduce fil_system.mutex contention a little. fil_node_t::complete_write(): Replaces fil_node_t::complete_io(). On both read and write completion, fil_space_t::release_for_io() will have to be called. fil_space_t::io(): Do not acquire fil_system.mutex in the normal code path. xb_delta_open_matching_space(): Do not try to open the system tablespace which was already opened. This fixes a file sharing violation in mariabackup --prepare --incremental. Reviewed by: Vladislav Vaintroub
2020-10-26 14:59:30 +01:00
space= nullptr;
}
if (!space)
buf_flush_discard_page(bpage);
else if (neighbors && space->is_rotational())
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
{
mysql_mutex_unlock(&buf_pool.mutex);
n->flushed+= buf_flush_try_neighbors(space, page_id, neighbors == 1,
true, n->flushed, max);
reacquire_mutex:
mysql_mutex_lock(&buf_pool.mutex);
}
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
else if (bpage->flush(true, space))
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
{
++n->flushed;
goto reacquire_mutex;
}
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
}
else
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
must_skip:
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
/* Can't evict or dispatch this block. Go to previous. */
ut_ad(buf_pool.lru_hp.is_hp(prev));
MDEV-28828 SIGSEGV in buf_flush_LRU_list_batch In commit 73fee39ea62037780c59161507e89dd76c10b7a3 (MDEV-27985) a regression was introduced that would cause bpage=nullptr to be referenced. buf_flush_LRU_list_batch(): Always terminate the loop upon encountering a null pointer.
2022-06-14 08:14:24 +02:00
retry:
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
bpage= buf_pool.lru_hp.get();
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
}
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
buf_pool.lru_hp.set(nullptr);
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
if (space)
MDEV-23855: Shrink fil_space_t Merge n_pending_ios, n_pending_ops to std::atomic<uint32_t> n_pending. Change some more fil_space_t members to uint32_t to reduce the memory footprint. fil_space_t::add(), fil_ibd_create(): Attach the already opened handle to the tablespace, and enforce the fil_system.n_open limit. dict_boot(): Initialize fil_system.max_assigned_id. srv_boot(): Call srv_thread_pool_init() before anything else, so that files should be opened in the correct mode on Windows. fil_ibd_create(): Create the file in OS_FILE_AIO mode, just like fil_node_open_file_low() does it. dict_table_t::is_accessible(): Replaces fil_table_accessible(). Reviewed by: Vladislav Vaintroub
2020-10-26 15:04:12 +01:00
space->release();
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
if (scanned)
MONITOR_INC_VALUE_CUMULATIVE(MONITOR_LRU_BATCH_SCANNED,
MONITOR_LRU_BATCH_SCANNED_NUM_CALL,
MONITOR_LRU_BATCH_SCANNED_PER_CALL,
scanned);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
}
MDEV-15058: Deprecate and ignore innodb_buffer_pool_instances Our benchmarking efforts indicate that the reasons for splitting the buf_pool in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 have mostly gone away, possibly as a result of mysql/mysql-server@ce6109ebfdedfdf185e391a0c97dc6d33867ed78 or similar work. Only in one write-heavy benchmark where the working set size is ten times the buffer pool size, the buf_pool->mutex would be less contended with 4 buffer pool instances than with 1 instance, in buf_page_io_complete(). That contention could be alleviated further by making more use of std::atomic and by splitting buf_pool_t::mutex further (MDEV-15053). We will deprecate and ignore the following parameters: innodb_buffer_pool_instances innodb_page_cleaners There will be only one buffer pool and one page cleaner task. In a number of INFORMATION_SCHEMA views, columns that indicated the buffer pool instance will be removed: information_schema.innodb_buffer_page.pool_id information_schema.innodb_buffer_page_lru.pool_id information_schema.innodb_buffer_pool_stats.pool_id information_schema.innodb_cmpmem.buffer_pool_instance information_schema.innodb_cmpmem_reset.buffer_pool_instance
2020-02-12 13:45:21 +01:00
/** Flush and move pages from LRU or unzip_LRU list to the free list.
Whether LRU or unzip_LRU is used depends on the state of the system.
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
@param max maximum number of blocks to make available in buf_pool.free
@return number of flushed pages */
static ulint buf_do_LRU_batch(ulint max)
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
{
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
const ulint n_unzip_LRU_evicted= buf_LRU_evict_from_unzip_LRU()
? buf_free_from_unzip_LRU_list_batch(max)
: 0;
flush_counters_t n;
n.flushed= 0;
n.evicted= n_unzip_LRU_evicted;
buf_flush_LRU_list_batch(max, &n);
MDEV-25085: Simplify instrumentation for LRU eviction Let us add the status variable innodb_buffer_pool_pages_LRU_freed to monitor the number of pages that were freed by a buffer pool LRU eviction scan, without flushing. Also, let us simplify the monitor interface: MONITOR_LRU_BATCH_FLUSH_COUNT, MONITOR_LRU_BATCH_FLUSH_PAGES, MONITOR_LRU_BATCH_EVICT_COUNT, MONITOR_LRU_BATCH_EVICT_PAGES: Remove. MONITOR_LRU_BATCH_FLUSH_TOTAL_PAGE: Track buf_lru_flush_page_count (innodb_buffer_pool_pages_LRU_flushed). MONITOR_LRU_BATCH_EVICT_TOTAL_PAGE: Track buf_lru_freed_page_count (buffer_pool_pages_LRU_freed). Reviewed by: Vladislav Vaintroub
2021-03-09 07:29:38 +01:00
mysql_mutex_assert_owner(&buf_pool.mutex);
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
if (const ulint evicted= n.evicted - n_unzip_LRU_evicted)
MDEV-25085: Simplify instrumentation for LRU eviction Let us add the status variable innodb_buffer_pool_pages_LRU_freed to monitor the number of pages that were freed by a buffer pool LRU eviction scan, without flushing. Also, let us simplify the monitor interface: MONITOR_LRU_BATCH_FLUSH_COUNT, MONITOR_LRU_BATCH_FLUSH_PAGES, MONITOR_LRU_BATCH_EVICT_COUNT, MONITOR_LRU_BATCH_EVICT_PAGES: Remove. MONITOR_LRU_BATCH_FLUSH_TOTAL_PAGE: Track buf_lru_flush_page_count (innodb_buffer_pool_pages_LRU_flushed). MONITOR_LRU_BATCH_EVICT_TOTAL_PAGE: Track buf_lru_freed_page_count (buffer_pool_pages_LRU_freed). Reviewed by: Vladislav Vaintroub
2021-03-09 07:29:38 +01:00
buf_lru_freed_page_count+= evicted;
if (n.flushed)
buf_lru_flush_page_count+= n.flushed;
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
return n.flushed;
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
}
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
/** This utility flushes dirty blocks from the end of the flush_list.
The calling thread is not allowed to own any latches on pages!
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
@param max_n maximum mumber of blocks to flush
@param lsn once an oldest_modification>=lsn is found, terminate the batch
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
@return number of blocks for which the write request was queued */
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
static ulint buf_do_flush_list_batch(ulint max_n, lsn_t lsn)
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
{
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
ulint count= 0;
ulint scanned= 0;
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
mysql_mutex_assert_owner(&buf_pool.mutex);
const auto neighbors= UT_LIST_GET_LEN(buf_pool.LRU) < BUF_LRU_OLD_MIN_LEN
? 0 : srv_flush_neighbors;
fil_space_t *space= nullptr;
MDEV-23855: Remove fil_system.LRU and reduce fil_system.mutex contention Also fixes MDEV-23929: innodb_flush_neighbors is not being ignored for system tablespace on SSD When the maximum configured number of file is exceeded, InnoDB will close data files. We used to maintain a fil_system.LRU list and a counter fil_node_t::n_pending to achieve this, at the huge cost of multiple fil_system.mutex operations per I/O operation. fil_node_open_file_low(): Implement a FIFO replacement policy: The last opened file will be moved to the end of fil_system.space_list, and files will be closed from the start of the list. However, we will not move tablespaces in fil_system.space_list while i_s_tablespaces_encryption_fill_table() is executing (producing output for INFORMATION_SCHEMA.INNODB_TABLESPACES_ENCRYPTION) because it may cause information of some tablespaces to go missing. We also avoid this in mariabackup --backup because datafiles_iter_next() assumes that the ordering is not changed. IORequest: Fold more parameters to IORequest::type. fil_space_t::io(): Replaces fil_io(). fil_space_t::flush(): Replaces fil_flush(). OS_AIO_IBUF: Remove. We will always issue synchronous reads of the change buffer pages in buf_read_page_low(). We will always ignore some errors for background reads. This should reduce fil_system.mutex contention a little. fil_node_t::complete_write(): Replaces fil_node_t::complete_io(). On both read and write completion, fil_space_t::release_for_io() will have to be called. fil_space_t::io(): Do not acquire fil_system.mutex in the normal code path. xb_delta_open_matching_space(): Do not try to open the system tablespace which was already opened. This fixes a file sharing violation in mariabackup --prepare --incremental. Reviewed by: Vladislav Vaintroub
2020-10-26 14:59:30 +01:00
uint32_t last_space_id= FIL_NULL;
static_assert(FIL_NULL > SRV_TMP_SPACE_ID, "consistency");
static_assert(FIL_NULL > SRV_SPACE_ID_UPPER_BOUND, "consistency");
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
/* Start from the end of the list looking for a suitable block to be
flushed. */
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
mysql_mutex_lock(&buf_pool.flush_list_mutex);
ulint len= UT_LIST_GET_LEN(buf_pool.flush_list);
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
for (buf_page_t *bpage= UT_LIST_GET_LAST(buf_pool.flush_list);
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
bpage && len && count < max_n; ++scanned, len--)
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
{
const lsn_t oldest_modification= bpage->oldest_modification();
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
if (oldest_modification >= lsn)
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
break;
MDEV-25801: buf_flush_dirty_pages() is very slow In commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) the implementation of buf_flush_dirty_pages() was replaced with a slow one, which would perform excessive scans of the buf_pool.flush_list and make little progress. buf_flush_list(), buf_flush_LRU(): Split from buf_flush_lists(). Vladislav Vaintroub noticed that we will not need to invoke log_flush_task.wait() for the LRU eviction flushing. buf_flush_list_space(): Replaces buf_flush_dirty_pages(). This is like buf_flush_list(), but operating on a single tablespace at a time. Writes at most innodb_io_capacity pages. Returns whether some of the tablespace might remain in the buffer pool.
2021-06-23 11:14:26 +02:00
ut_ad(bpage->in_file());
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
buf_page_t *prev= UT_LIST_GET_PREV(list, bpage);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
if (oldest_modification == 1)
MDEV-25801: buf_flush_dirty_pages() is very slow In commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) the implementation of buf_flush_dirty_pages() was replaced with a slow one, which would perform excessive scans of the buf_pool.flush_list and make little progress. buf_flush_list(), buf_flush_LRU(): Split from buf_flush_lists(). Vladislav Vaintroub noticed that we will not need to invoke log_flush_task.wait() for the LRU eviction flushing. buf_flush_list_space(): Replaces buf_flush_dirty_pages(). This is like buf_flush_list(), but operating on a single tablespace at a time. Writes at most innodb_io_capacity pages. Returns whether some of the tablespace might remain in the buffer pool.
2021-06-23 11:14:26 +02:00
{
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
buf_pool.delete_from_flush_list(bpage);
skip:
MDEV-25801: buf_flush_dirty_pages() is very slow In commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) the implementation of buf_flush_dirty_pages() was replaced with a slow one, which would perform excessive scans of the buf_pool.flush_list and make little progress. buf_flush_list(), buf_flush_LRU(): Split from buf_flush_lists(). Vladislav Vaintroub noticed that we will not need to invoke log_flush_task.wait() for the LRU eviction flushing. buf_flush_list_space(): Replaces buf_flush_dirty_pages(). This is like buf_flush_list(), but operating on a single tablespace at a time. Writes at most innodb_io_capacity pages. Returns whether some of the tablespace might remain in the buffer pool.
2021-06-23 11:14:26 +02:00
bpage= prev;
continue;
}
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
ut_ad(oldest_modification > 2);
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
if (!bpage->ready_for_flush())
goto skip;
MDEV-25801: buf_flush_dirty_pages() is very slow In commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) the implementation of buf_flush_dirty_pages() was replaced with a slow one, which would perform excessive scans of the buf_pool.flush_list and make little progress. buf_flush_list(), buf_flush_LRU(): Split from buf_flush_lists(). Vladislav Vaintroub noticed that we will not need to invoke log_flush_task.wait() for the LRU eviction flushing. buf_flush_list_space(): Replaces buf_flush_dirty_pages(). This is like buf_flush_list(), but operating on a single tablespace at a time. Writes at most innodb_io_capacity pages. Returns whether some of the tablespace might remain in the buffer pool.
2021-06-23 11:14:26 +02:00
/* In order not to degenerate this scan to O(n*n) we attempt to
preserve the pointer position. Any thread that would remove 'prev'
from buf_pool.flush_list must adjust the hazard pointer.
Note: A concurrent execution of buf_flush_list_space() may
terminate this scan prematurely. The buf_pool.n_flush_list()
should prevent multiple threads from executing
buf_do_flush_list_batch() concurrently,
but buf_flush_list_space() is ignoring that. */
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
buf_pool.flush_hp.set(prev);
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
mysql_mutex_unlock(&buf_pool.flush_list_mutex);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-25801: buf_flush_dirty_pages() is very slow In commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) the implementation of buf_flush_dirty_pages() was replaced with a slow one, which would perform excessive scans of the buf_pool.flush_list and make little progress. buf_flush_list(), buf_flush_LRU(): Split from buf_flush_lists(). Vladislav Vaintroub noticed that we will not need to invoke log_flush_task.wait() for the LRU eviction flushing. buf_flush_list_space(): Replaces buf_flush_dirty_pages(). This is like buf_flush_list(), but operating on a single tablespace at a time. Writes at most innodb_io_capacity pages. Returns whether some of the tablespace might remain in the buffer pool.
2021-06-23 11:14:26 +02:00
const page_id_t page_id(bpage->id());
const uint32_t space_id= page_id.space();
if (!space || space->id != space_id)
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
{
MDEV-25801: buf_flush_dirty_pages() is very slow In commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) the implementation of buf_flush_dirty_pages() was replaced with a slow one, which would perform excessive scans of the buf_pool.flush_list and make little progress. buf_flush_list(), buf_flush_LRU(): Split from buf_flush_lists(). Vladislav Vaintroub noticed that we will not need to invoke log_flush_task.wait() for the LRU eviction flushing. buf_flush_list_space(): Replaces buf_flush_dirty_pages(). This is like buf_flush_list(), but operating on a single tablespace at a time. Writes at most innodb_io_capacity pages. Returns whether some of the tablespace might remain in the buffer pool.
2021-06-23 11:14:26 +02:00
if (last_space_id != space_id)
MDEV-23855: Remove fil_system.LRU and reduce fil_system.mutex contention Also fixes MDEV-23929: innodb_flush_neighbors is not being ignored for system tablespace on SSD When the maximum configured number of file is exceeded, InnoDB will close data files. We used to maintain a fil_system.LRU list and a counter fil_node_t::n_pending to achieve this, at the huge cost of multiple fil_system.mutex operations per I/O operation. fil_node_open_file_low(): Implement a FIFO replacement policy: The last opened file will be moved to the end of fil_system.space_list, and files will be closed from the start of the list. However, we will not move tablespaces in fil_system.space_list while i_s_tablespaces_encryption_fill_table() is executing (producing output for INFORMATION_SCHEMA.INNODB_TABLESPACES_ENCRYPTION) because it may cause information of some tablespaces to go missing. We also avoid this in mariabackup --backup because datafiles_iter_next() assumes that the ordering is not changed. IORequest: Fold more parameters to IORequest::type. fil_space_t::io(): Replaces fil_io(). fil_space_t::flush(): Replaces fil_flush(). OS_AIO_IBUF: Remove. We will always issue synchronous reads of the change buffer pages in buf_read_page_low(). We will always ignore some errors for background reads. This should reduce fil_system.mutex contention a little. fil_node_t::complete_write(): Replaces fil_node_t::complete_io(). On both read and write completion, fil_space_t::release_for_io() will have to be called. fil_space_t::io(): Do not acquire fil_system.mutex in the normal code path. xb_delta_open_matching_space(): Do not try to open the system tablespace which was already opened. This fixes a file sharing violation in mariabackup --prepare --incremental. Reviewed by: Vladislav Vaintroub
2020-10-26 14:59:30 +01:00
{
MDEV-27985 buf_flush_freed_pages() causes InnoDB to hang buf_flush_freed_pages(): Assert that neither buf_pool.mutex nor buf_pool.flush_list_mutex are held. Simplify the loops. Return the tablespace and the number of pages written or punched. buf_flush_LRU_list_batch(), buf_do_flush_list_batch(): Release buf_pool.mutex before invoking buf_flush_space(). buf_flush_list_space(): Acquire the mutexes only after invoking buf_flush_freed_pages(). Reviewed by: Thirunarayanan Balathandayuthapani
2022-03-15 13:44:22 +01:00
mysql_mutex_lock(&buf_pool.flush_list_mutex);
buf_pool.flush_hp.set(bpage);
mysql_mutex_unlock(&buf_pool.flush_list_mutex);
mysql_mutex_unlock(&buf_pool.mutex);
MDEV-25801: buf_flush_dirty_pages() is very slow In commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) the implementation of buf_flush_dirty_pages() was replaced with a slow one, which would perform excessive scans of the buf_pool.flush_list and make little progress. buf_flush_list(), buf_flush_LRU(): Split from buf_flush_lists(). Vladislav Vaintroub noticed that we will not need to invoke log_flush_task.wait() for the LRU eviction flushing. buf_flush_list_space(): Replaces buf_flush_dirty_pages(). This is like buf_flush_list(), but operating on a single tablespace at a time. Writes at most innodb_io_capacity pages. Returns whether some of the tablespace might remain in the buffer pool.
2021-06-23 11:14:26 +02:00
if (space)
space->release();
MDEV-27985 buf_flush_freed_pages() causes InnoDB to hang buf_flush_freed_pages(): Assert that neither buf_pool.mutex nor buf_pool.flush_list_mutex are held. Simplify the loops. Return the tablespace and the number of pages written or punched. buf_flush_LRU_list_batch(), buf_do_flush_list_batch(): Release buf_pool.mutex before invoking buf_flush_space(). buf_flush_list_space(): Acquire the mutexes only after invoking buf_flush_freed_pages(). Reviewed by: Thirunarayanan Balathandayuthapani
2022-03-15 13:44:22 +01:00
auto p= buf_flush_space(space_id);
space= p.first;
MDEV-25801: buf_flush_dirty_pages() is very slow In commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) the implementation of buf_flush_dirty_pages() was replaced with a slow one, which would perform excessive scans of the buf_pool.flush_list and make little progress. buf_flush_list(), buf_flush_LRU(): Split from buf_flush_lists(). Vladislav Vaintroub noticed that we will not need to invoke log_flush_task.wait() for the LRU eviction flushing. buf_flush_list_space(): Replaces buf_flush_dirty_pages(). This is like buf_flush_list(), but operating on a single tablespace at a time. Writes at most innodb_io_capacity pages. Returns whether some of the tablespace might remain in the buffer pool.
2021-06-23 11:14:26 +02:00
last_space_id= space_id;
MDEV-27985 buf_flush_freed_pages() causes InnoDB to hang buf_flush_freed_pages(): Assert that neither buf_pool.mutex nor buf_pool.flush_list_mutex are held. Simplify the loops. Return the tablespace and the number of pages written or punched. buf_flush_LRU_list_batch(), buf_do_flush_list_batch(): Release buf_pool.mutex before invoking buf_flush_space(). buf_flush_list_space(): Acquire the mutexes only after invoking buf_flush_freed_pages(). Reviewed by: Thirunarayanan Balathandayuthapani
2022-03-15 13:44:22 +01:00
mysql_mutex_lock(&buf_pool.mutex);
if (p.second)
buf_pool.stat.n_pages_written+= p.second;
mysql_mutex_lock(&buf_pool.flush_list_mutex);
bpage= buf_pool.flush_hp.get();
if (!bpage)
break;
if (bpage->id() != page_id)
continue;
buf_pool.flush_hp.set(UT_LIST_GET_PREV(list, bpage));
if (bpage->oldest_modification() <= 1 || !bpage->ready_for_flush())
goto next;
mysql_mutex_unlock(&buf_pool.flush_list_mutex);
MDEV-23855: Remove fil_system.LRU and reduce fil_system.mutex contention Also fixes MDEV-23929: innodb_flush_neighbors is not being ignored for system tablespace on SSD When the maximum configured number of file is exceeded, InnoDB will close data files. We used to maintain a fil_system.LRU list and a counter fil_node_t::n_pending to achieve this, at the huge cost of multiple fil_system.mutex operations per I/O operation. fil_node_open_file_low(): Implement a FIFO replacement policy: The last opened file will be moved to the end of fil_system.space_list, and files will be closed from the start of the list. However, we will not move tablespaces in fil_system.space_list while i_s_tablespaces_encryption_fill_table() is executing (producing output for INFORMATION_SCHEMA.INNODB_TABLESPACES_ENCRYPTION) because it may cause information of some tablespaces to go missing. We also avoid this in mariabackup --backup because datafiles_iter_next() assumes that the ordering is not changed. IORequest: Fold more parameters to IORequest::type. fil_space_t::io(): Replaces fil_io(). fil_space_t::flush(): Replaces fil_flush(). OS_AIO_IBUF: Remove. We will always issue synchronous reads of the change buffer pages in buf_read_page_low(). We will always ignore some errors for background reads. This should reduce fil_system.mutex contention a little. fil_node_t::complete_write(): Replaces fil_node_t::complete_io(). On both read and write completion, fil_space_t::release_for_io() will have to be called. fil_space_t::io(): Do not acquire fil_system.mutex in the normal code path. xb_delta_open_matching_space(): Do not try to open the system tablespace which was already opened. This fixes a file sharing violation in mariabackup --prepare --incremental. Reviewed by: Vladislav Vaintroub
2020-10-26 14:59:30 +01:00
}
MDEV-25801: buf_flush_dirty_pages() is very slow In commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) the implementation of buf_flush_dirty_pages() was replaced with a slow one, which would perform excessive scans of the buf_pool.flush_list and make little progress. buf_flush_list(), buf_flush_LRU(): Split from buf_flush_lists(). Vladislav Vaintroub noticed that we will not need to invoke log_flush_task.wait() for the LRU eviction flushing. buf_flush_list_space(): Replaces buf_flush_dirty_pages(). This is like buf_flush_list(), but operating on a single tablespace at a time. Writes at most innodb_io_capacity pages. Returns whether some of the tablespace might remain in the buffer pool.
2021-06-23 11:14:26 +02:00
else
ut_ad(!space);
}
else if (space->is_stopping())
{
space->release();
space= nullptr;
}
MDEV-23855: Remove fil_system.LRU and reduce fil_system.mutex contention Also fixes MDEV-23929: innodb_flush_neighbors is not being ignored for system tablespace on SSD When the maximum configured number of file is exceeded, InnoDB will close data files. We used to maintain a fil_system.LRU list and a counter fil_node_t::n_pending to achieve this, at the huge cost of multiple fil_system.mutex operations per I/O operation. fil_node_open_file_low(): Implement a FIFO replacement policy: The last opened file will be moved to the end of fil_system.space_list, and files will be closed from the start of the list. However, we will not move tablespaces in fil_system.space_list while i_s_tablespaces_encryption_fill_table() is executing (producing output for INFORMATION_SCHEMA.INNODB_TABLESPACES_ENCRYPTION) because it may cause information of some tablespaces to go missing. We also avoid this in mariabackup --backup because datafiles_iter_next() assumes that the ordering is not changed. IORequest: Fold more parameters to IORequest::type. fil_space_t::io(): Replaces fil_io(). fil_space_t::flush(): Replaces fil_flush(). OS_AIO_IBUF: Remove. We will always issue synchronous reads of the change buffer pages in buf_read_page_low(). We will always ignore some errors for background reads. This should reduce fil_system.mutex contention a little. fil_node_t::complete_write(): Replaces fil_node_t::complete_io(). On both read and write completion, fil_space_t::release_for_io() will have to be called. fil_space_t::io(): Do not acquire fil_system.mutex in the normal code path. xb_delta_open_matching_space(): Do not try to open the system tablespace which was already opened. This fixes a file sharing violation in mariabackup --prepare --incremental. Reviewed by: Vladislav Vaintroub
2020-10-26 14:59:30 +01:00
MDEV-25801: buf_flush_dirty_pages() is very slow In commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) the implementation of buf_flush_dirty_pages() was replaced with a slow one, which would perform excessive scans of the buf_pool.flush_list and make little progress. buf_flush_list(), buf_flush_LRU(): Split from buf_flush_lists(). Vladislav Vaintroub noticed that we will not need to invoke log_flush_task.wait() for the LRU eviction flushing. buf_flush_list_space(): Replaces buf_flush_dirty_pages(). This is like buf_flush_list(), but operating on a single tablespace at a time. Writes at most innodb_io_capacity pages. Returns whether some of the tablespace might remain in the buffer pool.
2021-06-23 11:14:26 +02:00
if (!space)
buf_flush_discard_page(bpage);
else if (neighbors && space->is_rotational())
{
mysql_mutex_unlock(&buf_pool.mutex);
count+= buf_flush_try_neighbors(space, page_id, neighbors == 1,
false, count, max_n);
reacquire_mutex:
mysql_mutex_lock(&buf_pool.mutex);
}
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
else if (bpage->flush(false, space))
MDEV-25801: buf_flush_dirty_pages() is very slow In commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) the implementation of buf_flush_dirty_pages() was replaced with a slow one, which would perform excessive scans of the buf_pool.flush_list and make little progress. buf_flush_list(), buf_flush_LRU(): Split from buf_flush_lists(). Vladislav Vaintroub noticed that we will not need to invoke log_flush_task.wait() for the LRU eviction flushing. buf_flush_list_space(): Replaces buf_flush_dirty_pages(). This is like buf_flush_list(), but operating on a single tablespace at a time. Writes at most innodb_io_capacity pages. Returns whether some of the tablespace might remain in the buffer pool.
2021-06-23 11:14:26 +02:00
{
++count;
goto reacquire_mutex;
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
}
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
mysql_mutex_lock(&buf_pool.flush_list_mutex);
MDEV-27985 buf_flush_freed_pages() causes InnoDB to hang buf_flush_freed_pages(): Assert that neither buf_pool.mutex nor buf_pool.flush_list_mutex are held. Simplify the loops. Return the tablespace and the number of pages written or punched. buf_flush_LRU_list_batch(), buf_do_flush_list_batch(): Release buf_pool.mutex before invoking buf_flush_space(). buf_flush_list_space(): Acquire the mutexes only after invoking buf_flush_freed_pages(). Reviewed by: Thirunarayanan Balathandayuthapani
2022-03-15 13:44:22 +01:00
next:
MDEV-25801: buf_flush_dirty_pages() is very slow In commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) the implementation of buf_flush_dirty_pages() was replaced with a slow one, which would perform excessive scans of the buf_pool.flush_list and make little progress. buf_flush_list(), buf_flush_LRU(): Split from buf_flush_lists(). Vladislav Vaintroub noticed that we will not need to invoke log_flush_task.wait() for the LRU eviction flushing. buf_flush_list_space(): Replaces buf_flush_dirty_pages(). This is like buf_flush_list(), but operating on a single tablespace at a time. Writes at most innodb_io_capacity pages. Returns whether some of the tablespace might remain in the buffer pool.
2021-06-23 11:14:26 +02:00
bpage= buf_pool.flush_hp.get();
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
}
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
buf_pool.flush_hp.set(nullptr);
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
mysql_mutex_unlock(&buf_pool.flush_list_mutex);
if (space)
MDEV-23855: Shrink fil_space_t Merge n_pending_ios, n_pending_ops to std::atomic<uint32_t> n_pending. Change some more fil_space_t members to uint32_t to reduce the memory footprint. fil_space_t::add(), fil_ibd_create(): Attach the already opened handle to the tablespace, and enforce the fil_system.n_open limit. dict_boot(): Initialize fil_system.max_assigned_id. srv_boot(): Call srv_thread_pool_init() before anything else, so that files should be opened in the correct mode on Windows. fil_ibd_create(): Create the file in OS_FILE_AIO mode, just like fil_node_open_file_low() does it. dict_table_t::is_accessible(): Replaces fil_table_accessible(). Reviewed by: Vladislav Vaintroub
2020-10-26 15:04:12 +01:00
space->release();
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
if (scanned)
MONITOR_INC_VALUE_CUMULATIVE(MONITOR_FLUSH_BATCH_SCANNED,
MONITOR_FLUSH_BATCH_SCANNED_NUM_CALL,
MONITOR_FLUSH_BATCH_SCANNED_PER_CALL,
scanned);
if (count)
MONITOR_INC_VALUE_CUMULATIVE(MONITOR_FLUSH_BATCH_TOTAL_PAGE,
MONITOR_FLUSH_BATCH_COUNT,
MONITOR_FLUSH_BATCH_PAGES,
count);
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
mysql_mutex_assert_owner(&buf_pool.mutex);
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
return count;
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
}
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
/** Wait until a flush batch ends.
@param lru true=buf_pool.LRU; false=buf_pool.flush_list */
void buf_flush_wait_batch_end(bool lru)
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
{
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
const auto &n_flush= lru ? buf_pool.n_flush_LRU_ : buf_pool.n_flush_list_;
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
if (n_flush)
{
auto cond= lru ? &buf_pool.done_flush_LRU : &buf_pool.done_flush_list;
tpool::tpool_wait_begin();
thd_wait_begin(nullptr, THD_WAIT_DISKIO);
do
MDEV-23399 fixup: Use plain pthread_cond The condition variables that were introduced in commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) are never instrumented with PERFORMANCE_SCHEMA. Let us avoid the storage overhead and dead code.
2021-02-07 11:19:24 +01:00
my_cond_wait(cond, &buf_pool.mutex.m_mutex);
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
while (n_flush);
tpool::tpool_wait_end();
thd_wait_end(nullptr);
MDEV-23399 fixup: Use plain pthread_cond The condition variables that were introduced in commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) are never instrumented with PERFORMANCE_SCHEMA. Let us avoid the storage overhead and dead code.
2021-02-07 11:19:24 +01:00
pthread_cond_broadcast(cond);
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
}
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
}
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
/** Write out dirty blocks from buf_pool.flush_list.
@param max_n wished maximum mumber of blocks flushed
MDEV-25801: buf_flush_dirty_pages() is very slow In commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) the implementation of buf_flush_dirty_pages() was replaced with a slow one, which would perform excessive scans of the buf_pool.flush_list and make little progress. buf_flush_list(), buf_flush_LRU(): Split from buf_flush_lists(). Vladislav Vaintroub noticed that we will not need to invoke log_flush_task.wait() for the LRU eviction flushing. buf_flush_list_space(): Replaces buf_flush_dirty_pages(). This is like buf_flush_list(), but operating on a single tablespace at a time. Writes at most innodb_io_capacity pages. Returns whether some of the tablespace might remain in the buffer pool.
2021-06-23 11:14:26 +02:00
@param lsn buf_pool.get_oldest_modification(LSN_MAX) target
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
@return the number of processed pages
MDEV-25801: buf_flush_dirty_pages() is very slow In commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) the implementation of buf_flush_dirty_pages() was replaced with a slow one, which would perform excessive scans of the buf_pool.flush_list and make little progress. buf_flush_list(), buf_flush_LRU(): Split from buf_flush_lists(). Vladislav Vaintroub noticed that we will not need to invoke log_flush_task.wait() for the LRU eviction flushing. buf_flush_list_space(): Replaces buf_flush_dirty_pages(). This is like buf_flush_list(), but operating on a single tablespace at a time. Writes at most innodb_io_capacity pages. Returns whether some of the tablespace might remain in the buffer pool.
2021-06-23 11:14:26 +02:00
@retval 0 if a buf_pool.flush_list batch is already running */
MDEV-27416 InnoDB hang in buf_flush_wait_flushed(), on log checkpoint InnoDB could sometimes hang when triggering a log checkpoint. This is due to commit 7b1252c03d7131754d9503560fe507b33ca1f8b4 (MDEV-24278), which introduced an untimed wait to buf_flush_page_cleaner(). The hang was noticed by occasional failures of IMPORT TABLESPACE tests, such as innodb.innodb-wl5522, which would (unnecessarily) invoke log_make_checkpoint() from row_import_cleanup(). The reason of the hang was that buf_flush_page_cleaner() would enter untimed sleep despite buf_flush_sync_lsn being set. The exact failure scenario is unclear, because buf_flush_sync_lsn should actually be protected by buf_pool.flush_list_mutex. We prevent the hang by invoking buf_pool.page_cleaner_set_idle(false) whenever we are setting buf_flush_sync_lsn and signaling buf_pool.do_flush_list. The bulk of these changes was originally developed as a preparation for MDEV-26827, to invoke buf_flush_list() from fewer threads, and tested on 10.6 by Matthias Leich. This fix was tested by running 100 repetitions of 100 concurrent instances of the test innodb.innodb-wl5522 on a RelWithDebInfo build, using ext4fs and innodb_flush_method=O_DIRECT on a SATA SSD with 4096-byte block size. During the test, the call to log_make_checkpoint() in row_import_cleanup() was present. buf_flush_list(): Make static. buf_flush_wait(): Wait for buf_pool.get_oldest_modification() to reach a target, by work done in the buf_flush_page_cleaner. If buf_flush_sync_lsn is going to be set, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_ahead(): If buf_flush_sync_lsn or buf_flush_async_lsn is going to be set and the page cleaner woken up, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_wait_flushed(): Invoke buf_flush_wait(). buf_flush_sync(): Invoke recv_sys.apply() at the start in case crash recovery is active. Invoke buf_flush_wait(). buf_flush_sync_batch(): A lower-level variant of buf_flush_sync() that is only called by recv_sys_t::apply(). buf_flush_sync_for_checkpoint(): Do not trigger log apply or checkpoint during recovery. buf_dblwr_t::create(): Only initiate a buffer pool flush, not a checkpoint. row_import_cleanup(): Do not unnecessarily invoke log_make_checkpoint(). Invoking buf_flush_list_space() before starting to generate redo log for the imported tablespace should suffice. srv_prepare_to_delete_redo_log_file(): Set recv_sys.recovery_on in order to prevent buf_flush_sync_for_checkpoint() from initiating a checkpoint while the log is inaccessible. Remove a wait loop that is already part of buf_flush_sync(). Do not invoke fil_names_clear() if the log is being upgraded, because the FILE_MODIFY record is specific to the latest format. create_log_file(): Clear recv_sys.recovery_on only after calling log_make_checkpoint(), to prevent buf_flush_page_cleaner from invoking a checkpoint. innodb_shutdown(): Simplify the logic in mariadb-backup --prepare. os_aio_wait_until_no_pending_writes(): Update the function comment. Apart from row_quiesce_table_start() during FLUSH TABLES...FOR EXPORT, this is being called by buf_flush_list_space(), which is invoked by ALTER TABLE...IMPORT TABLESPACE as well as some encryption operations.
2022-01-04 06:40:31 +01:00
static ulint buf_flush_list(ulint max_n= ULINT_UNDEFINED, lsn_t lsn= LSN_MAX)
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
{
MDEV-25801: buf_flush_dirty_pages() is very slow In commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) the implementation of buf_flush_dirty_pages() was replaced with a slow one, which would perform excessive scans of the buf_pool.flush_list and make little progress. buf_flush_list(), buf_flush_LRU(): Split from buf_flush_lists(). Vladislav Vaintroub noticed that we will not need to invoke log_flush_task.wait() for the LRU eviction flushing. buf_flush_list_space(): Replaces buf_flush_dirty_pages(). This is like buf_flush_list(), but operating on a single tablespace at a time. Writes at most innodb_io_capacity pages. Returns whether some of the tablespace might remain in the buffer pool.
2021-06-23 11:14:26 +02:00
ut_ad(lsn);
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
if (buf_pool.n_flush_list())
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
return 0;
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
mysql_mutex_lock(&buf_pool.mutex);
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
const bool running= buf_pool.n_flush_list_ != 0;
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
/* FIXME: we are performing a dirty read of buf_pool.flush_list.count
while not holding buf_pool.flush_list_mutex */
MDEV-25801: buf_flush_dirty_pages() is very slow In commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) the implementation of buf_flush_dirty_pages() was replaced with a slow one, which would perform excessive scans of the buf_pool.flush_list and make little progress. buf_flush_list(), buf_flush_LRU(): Split from buf_flush_lists(). Vladislav Vaintroub noticed that we will not need to invoke log_flush_task.wait() for the LRU eviction flushing. buf_flush_list_space(): Replaces buf_flush_dirty_pages(). This is like buf_flush_list(), but operating on a single tablespace at a time. Writes at most innodb_io_capacity pages. Returns whether some of the tablespace might remain in the buffer pool.
2021-06-23 11:14:26 +02:00
if (running || !UT_LIST_GET_LEN(buf_pool.flush_list))
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
{
MDEV-23855: Remove fil_system.LRU and reduce fil_system.mutex contention Also fixes MDEV-23929: innodb_flush_neighbors is not being ignored for system tablespace on SSD When the maximum configured number of file is exceeded, InnoDB will close data files. We used to maintain a fil_system.LRU list and a counter fil_node_t::n_pending to achieve this, at the huge cost of multiple fil_system.mutex operations per I/O operation. fil_node_open_file_low(): Implement a FIFO replacement policy: The last opened file will be moved to the end of fil_system.space_list, and files will be closed from the start of the list. However, we will not move tablespaces in fil_system.space_list while i_s_tablespaces_encryption_fill_table() is executing (producing output for INFORMATION_SCHEMA.INNODB_TABLESPACES_ENCRYPTION) because it may cause information of some tablespaces to go missing. We also avoid this in mariabackup --backup because datafiles_iter_next() assumes that the ordering is not changed. IORequest: Fold more parameters to IORequest::type. fil_space_t::io(): Replaces fil_io(). fil_space_t::flush(): Replaces fil_flush(). OS_AIO_IBUF: Remove. We will always issue synchronous reads of the change buffer pages in buf_read_page_low(). We will always ignore some errors for background reads. This should reduce fil_system.mutex contention a little. fil_node_t::complete_write(): Replaces fil_node_t::complete_io(). On both read and write completion, fil_space_t::release_for_io() will have to be called. fil_space_t::io(): Do not acquire fil_system.mutex in the normal code path. xb_delta_open_matching_space(): Do not try to open the system tablespace which was already opened. This fixes a file sharing violation in mariabackup --prepare --incremental. Reviewed by: Vladislav Vaintroub
2020-10-26 14:59:30 +01:00
if (!running)
MDEV-25801: buf_flush_dirty_pages() is very slow In commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) the implementation of buf_flush_dirty_pages() was replaced with a slow one, which would perform excessive scans of the buf_pool.flush_list and make little progress. buf_flush_list(), buf_flush_LRU(): Split from buf_flush_lists(). Vladislav Vaintroub noticed that we will not need to invoke log_flush_task.wait() for the LRU eviction flushing. buf_flush_list_space(): Replaces buf_flush_dirty_pages(). This is like buf_flush_list(), but operating on a single tablespace at a time. Writes at most innodb_io_capacity pages. Returns whether some of the tablespace might remain in the buffer pool.
2021-06-23 11:14:26 +02:00
pthread_cond_broadcast(&buf_pool.done_flush_list);
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
mysql_mutex_unlock(&buf_pool.mutex);
return 0;
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
}
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
buf_pool.n_flush_list_++;
const ulint n_flushed= buf_do_flush_list_batch(max_n, lsn);
const ulint n_flushing= --buf_pool.n_flush_list_;
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
buf_pool.try_LRU_scan= true;
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
mysql_mutex_unlock(&buf_pool.mutex);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
if (!n_flushing)
MDEV-25801: buf_flush_dirty_pages() is very slow In commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) the implementation of buf_flush_dirty_pages() was replaced with a slow one, which would perform excessive scans of the buf_pool.flush_list and make little progress. buf_flush_list(), buf_flush_LRU(): Split from buf_flush_lists(). Vladislav Vaintroub noticed that we will not need to invoke log_flush_task.wait() for the LRU eviction flushing. buf_flush_list_space(): Replaces buf_flush_dirty_pages(). This is like buf_flush_list(), but operating on a single tablespace at a time. Writes at most innodb_io_capacity pages. Returns whether some of the tablespace might remain in the buffer pool.
2021-06-23 11:14:26 +02:00
pthread_cond_broadcast(&buf_pool.done_flush_list);
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
buf_dblwr.flush_buffered_writes();
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-25801: buf_flush_dirty_pages() is very slow In commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) the implementation of buf_flush_dirty_pages() was replaced with a slow one, which would perform excessive scans of the buf_pool.flush_list and make little progress. buf_flush_list(), buf_flush_LRU(): Split from buf_flush_lists(). Vladislav Vaintroub noticed that we will not need to invoke log_flush_task.wait() for the LRU eviction flushing. buf_flush_list_space(): Replaces buf_flush_dirty_pages(). This is like buf_flush_list(), but operating on a single tablespace at a time. Writes at most innodb_io_capacity pages. Returns whether some of the tablespace might remain in the buffer pool.
2021-06-23 11:14:26 +02:00
DBUG_PRINT("ib_buf", ("flush_list completed, " ULINTPF " pages", n_flushed));
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
return n_flushed;
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
}
MDEV-25801: buf_flush_dirty_pages() is very slow In commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) the implementation of buf_flush_dirty_pages() was replaced with a slow one, which would perform excessive scans of the buf_pool.flush_list and make little progress. buf_flush_list(), buf_flush_LRU(): Split from buf_flush_lists(). Vladislav Vaintroub noticed that we will not need to invoke log_flush_task.wait() for the LRU eviction flushing. buf_flush_list_space(): Replaces buf_flush_dirty_pages(). This is like buf_flush_list(), but operating on a single tablespace at a time. Writes at most innodb_io_capacity pages. Returns whether some of the tablespace might remain in the buffer pool.
2021-06-23 11:14:26 +02:00
/** Try to flush all the dirty pages that belong to a given tablespace.
@param space tablespace
@param n_flushed number of pages written
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
@return whether the flush for some pages might not have been initiated */
MDEV-25801: buf_flush_dirty_pages() is very slow In commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) the implementation of buf_flush_dirty_pages() was replaced with a slow one, which would perform excessive scans of the buf_pool.flush_list and make little progress. buf_flush_list(), buf_flush_LRU(): Split from buf_flush_lists(). Vladislav Vaintroub noticed that we will not need to invoke log_flush_task.wait() for the LRU eviction flushing. buf_flush_list_space(): Replaces buf_flush_dirty_pages(). This is like buf_flush_list(), but operating on a single tablespace at a time. Writes at most innodb_io_capacity pages. Returns whether some of the tablespace might remain in the buffer pool.
2021-06-23 11:14:26 +02:00
bool buf_flush_list_space(fil_space_t *space, ulint *n_flushed)
{
const auto space_id= space->id;
ut_ad(space_id <= SRV_SPACE_ID_UPPER_BOUND);
bool may_have_skipped= false;
ulint max_n_flush= srv_io_capacity;
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-25801: buf_flush_dirty_pages() is very slow In commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) the implementation of buf_flush_dirty_pages() was replaced with a slow one, which would perform excessive scans of the buf_pool.flush_list and make little progress. buf_flush_list(), buf_flush_LRU(): Split from buf_flush_lists(). Vladislav Vaintroub noticed that we will not need to invoke log_flush_task.wait() for the LRU eviction flushing. buf_flush_list_space(): Replaces buf_flush_dirty_pages(). This is like buf_flush_list(), but operating on a single tablespace at a time. Writes at most innodb_io_capacity pages. Returns whether some of the tablespace might remain in the buffer pool.
2021-06-23 11:14:26 +02:00
bool acquired= space->acquire();
MDEV-27985 buf_flush_freed_pages() causes InnoDB to hang buf_flush_freed_pages(): Assert that neither buf_pool.mutex nor buf_pool.flush_list_mutex are held. Simplify the loops. Return the tablespace and the number of pages written or punched. buf_flush_LRU_list_batch(), buf_do_flush_list_batch(): Release buf_pool.mutex before invoking buf_flush_space(). buf_flush_list_space(): Acquire the mutexes only after invoking buf_flush_freed_pages(). Reviewed by: Thirunarayanan Balathandayuthapani
2022-03-15 13:44:22 +01:00
{
Cleanup: Make fil_space_t::freed_ranges private fil_space_t::is_freed(): Check if a page is in freed_ranges. fil_space_t::flush_freed(): Replaces buf_flush_freed_pages().
2022-06-06 10:55:29 +02:00
const uint32_t written{space->flush_freed(acquired)};
MDEV-27985 buf_flush_freed_pages() causes InnoDB to hang buf_flush_freed_pages(): Assert that neither buf_pool.mutex nor buf_pool.flush_list_mutex are held. Simplify the loops. Return the tablespace and the number of pages written or punched. buf_flush_LRU_list_batch(), buf_do_flush_list_batch(): Release buf_pool.mutex before invoking buf_flush_space(). buf_flush_list_space(): Acquire the mutexes only after invoking buf_flush_freed_pages(). Reviewed by: Thirunarayanan Balathandayuthapani
2022-03-15 13:44:22 +01:00
mysql_mutex_lock(&buf_pool.mutex);
if (written)
buf_pool.stat.n_pages_written+= written;
}
mysql_mutex_lock(&buf_pool.flush_list_mutex);
MDEV-25801: buf_flush_dirty_pages() is very slow In commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) the implementation of buf_flush_dirty_pages() was replaced with a slow one, which would perform excessive scans of the buf_pool.flush_list and make little progress. buf_flush_list(), buf_flush_LRU(): Split from buf_flush_lists(). Vladislav Vaintroub noticed that we will not need to invoke log_flush_task.wait() for the LRU eviction flushing. buf_flush_list_space(): Replaces buf_flush_dirty_pages(). This is like buf_flush_list(), but operating on a single tablespace at a time. Writes at most innodb_io_capacity pages. Returns whether some of the tablespace might remain in the buffer pool.
2021-06-23 11:14:26 +02:00
for (buf_page_t *bpage= UT_LIST_GET_LAST(buf_pool.flush_list); bpage; )
{
ut_ad(bpage->oldest_modification());
ut_ad(bpage->in_file());
buf_page_t *prev= UT_LIST_GET_PREV(list, bpage);
if (bpage->id().space() != space_id);
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
else if (bpage->oldest_modification() == 1)
buf_pool.delete_from_flush_list(bpage);
MDEV-25801: buf_flush_dirty_pages() is very slow In commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) the implementation of buf_flush_dirty_pages() was replaced with a slow one, which would perform excessive scans of the buf_pool.flush_list and make little progress. buf_flush_list(), buf_flush_LRU(): Split from buf_flush_lists(). Vladislav Vaintroub noticed that we will not need to invoke log_flush_task.wait() for the LRU eviction flushing. buf_flush_list_space(): Replaces buf_flush_dirty_pages(). This is like buf_flush_list(), but operating on a single tablespace at a time. Writes at most innodb_io_capacity pages. Returns whether some of the tablespace might remain in the buffer pool.
2021-06-23 11:14:26 +02:00
else if (!bpage->ready_for_flush())
may_have_skipped= true;
else
{
/* In order not to degenerate this scan to O(n*n) we attempt to
preserve the pointer position. Any thread that would remove 'prev'
from buf_pool.flush_list must adjust the hazard pointer.
Note: Multiple executions of buf_flush_list_space() may be
interleaved, and also buf_do_flush_list_batch() may be running
concurrently. This may terminate our iteration prematurely,
leading us to return may_have_skipped=true. */
buf_pool.flush_hp.set(prev);
mysql_mutex_unlock(&buf_pool.flush_list_mutex);
if (!acquired)
{
was_freed:
buf_flush_discard_page(bpage);
}
else
{
if (space->is_stopping())
{
space->release();
acquired= false;
goto was_freed;
}
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
if (!bpage->flush(false, space))
MDEV-25801: buf_flush_dirty_pages() is very slow In commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) the implementation of buf_flush_dirty_pages() was replaced with a slow one, which would perform excessive scans of the buf_pool.flush_list and make little progress. buf_flush_list(), buf_flush_LRU(): Split from buf_flush_lists(). Vladislav Vaintroub noticed that we will not need to invoke log_flush_task.wait() for the LRU eviction flushing. buf_flush_list_space(): Replaces buf_flush_dirty_pages(). This is like buf_flush_list(), but operating on a single tablespace at a time. Writes at most innodb_io_capacity pages. Returns whether some of the tablespace might remain in the buffer pool.
2021-06-23 11:14:26 +02:00
{
may_have_skipped= true;
mysql_mutex_lock(&buf_pool.flush_list_mutex);
goto next_after_skip;
}
if (n_flushed)
++*n_flushed;
if (!--max_n_flush)
{
mysql_mutex_lock(&buf_pool.mutex);
mysql_mutex_lock(&buf_pool.flush_list_mutex);
may_have_skipped= true;
break;
}
mysql_mutex_lock(&buf_pool.mutex);
}
mysql_mutex_lock(&buf_pool.flush_list_mutex);
if (!buf_pool.flush_hp.is_hp(prev))
may_have_skipped= true;
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
next_after_skip:
MDEV-25801: buf_flush_dirty_pages() is very slow In commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) the implementation of buf_flush_dirty_pages() was replaced with a slow one, which would perform excessive scans of the buf_pool.flush_list and make little progress. buf_flush_list(), buf_flush_LRU(): Split from buf_flush_lists(). Vladislav Vaintroub noticed that we will not need to invoke log_flush_task.wait() for the LRU eviction flushing. buf_flush_list_space(): Replaces buf_flush_dirty_pages(). This is like buf_flush_list(), but operating on a single tablespace at a time. Writes at most innodb_io_capacity pages. Returns whether some of the tablespace might remain in the buffer pool.
2021-06-23 11:14:26 +02:00
bpage= buf_pool.flush_hp.get();
continue;
}
bpage= prev;
}
/* Note: this loop may have been executed concurrently with
buf_do_flush_list_batch() as well as other threads executing
buf_flush_list_space(). We should always return true from
buf_flush_list_space() if that should be the case; in
buf_do_flush_list_batch() we will simply perform less work. */
buf_pool.flush_hp.set(nullptr);
mysql_mutex_unlock(&buf_pool.flush_list_mutex);
buf_pool.try_LRU_scan= true;
mysql_mutex_unlock(&buf_pool.mutex);
if (acquired)
space->release();
if (space->purpose == FIL_TYPE_IMPORT)
os_aio_wait_until_no_pending_writes();
else
buf_dblwr.flush_buffered_writes();
return may_have_skipped;
}
/** Write out dirty blocks from buf_pool.LRU.
@param max_n wished maximum mumber of blocks flushed
@return the number of processed pages
@retval 0 if a buf_pool.LRU batch is already running */
ulint buf_flush_LRU(ulint max_n)
{
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
if (buf_pool.n_flush_LRU())
MDEV-25801: buf_flush_dirty_pages() is very slow In commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) the implementation of buf_flush_dirty_pages() was replaced with a slow one, which would perform excessive scans of the buf_pool.flush_list and make little progress. buf_flush_list(), buf_flush_LRU(): Split from buf_flush_lists(). Vladislav Vaintroub noticed that we will not need to invoke log_flush_task.wait() for the LRU eviction flushing. buf_flush_list_space(): Replaces buf_flush_dirty_pages(). This is like buf_flush_list(), but operating on a single tablespace at a time. Writes at most innodb_io_capacity pages. Returns whether some of the tablespace might remain in the buffer pool.
2021-06-23 11:14:26 +02:00
return 0;
MDEV-26166: non-functional changes
2021-07-20 09:55:03 +02:00
log_buffer_flush_to_disk();
MDEV-25801: buf_flush_dirty_pages() is very slow In commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) the implementation of buf_flush_dirty_pages() was replaced with a slow one, which would perform excessive scans of the buf_pool.flush_list and make little progress. buf_flush_list(), buf_flush_LRU(): Split from buf_flush_lists(). Vladislav Vaintroub noticed that we will not need to invoke log_flush_task.wait() for the LRU eviction flushing. buf_flush_list_space(): Replaces buf_flush_dirty_pages(). This is like buf_flush_list(), but operating on a single tablespace at a time. Writes at most innodb_io_capacity pages. Returns whether some of the tablespace might remain in the buffer pool.
2021-06-23 11:14:26 +02:00
mysql_mutex_lock(&buf_pool.mutex);
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
if (buf_pool.n_flush_LRU_)
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
{
mysql_mutex_unlock(&buf_pool.mutex);
return 0;
}
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
buf_pool.n_flush_LRU_++;
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-25801: buf_flush_dirty_pages() is very slow In commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) the implementation of buf_flush_dirty_pages() was replaced with a slow one, which would perform excessive scans of the buf_pool.flush_list and make little progress. buf_flush_list(), buf_flush_LRU(): Split from buf_flush_lists(). Vladislav Vaintroub noticed that we will not need to invoke log_flush_task.wait() for the LRU eviction flushing. buf_flush_list_space(): Replaces buf_flush_dirty_pages(). This is like buf_flush_list(), but operating on a single tablespace at a time. Writes at most innodb_io_capacity pages. Returns whether some of the tablespace might remain in the buffer pool.
2021-06-23 11:14:26 +02:00
ulint n_flushed= buf_do_LRU_batch(max_n);
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
const ulint n_flushing= --buf_pool.n_flush_LRU_;
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
buf_pool.try_LRU_scan= true;
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
mysql_mutex_unlock(&buf_pool.mutex);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
if (!n_flushing)
MDEV-26004 Excessive wait times in buf_LRU_get_free_block() buf_LRU_get_free_block(): Initially wait for a single block to be freed, signaled by buf_pool.done_free. Only if that fails and no LRU eviction flushing batch is already running, we initiate a flushing batch that should serve all threads that are currently waiting in buf_LRU_get_free_block(). Note: In an extreme case, this may introduce a performance regression at larger numbers of connections. We observed this in sysbench oltp_update_index with 512MiB buffer pool, 4GiB of data on fast NVMe, and 1000 concurrent connections, on a 20-thread CPU. The contention point appears to be buf_pool.mutex, and the improvement would turn into a regression somewhere beyond 32 concurrent connections. On slower storage, such regression was not observed; instead, the throughput was improving and maximum latency was reduced. The excessive waits were pointed out by Vladislav Vaintroub.
2021-06-24 10:01:18 +02:00
{
MDEV-25801: buf_flush_dirty_pages() is very slow In commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) the implementation of buf_flush_dirty_pages() was replaced with a slow one, which would perform excessive scans of the buf_pool.flush_list and make little progress. buf_flush_list(), buf_flush_LRU(): Split from buf_flush_lists(). Vladislav Vaintroub noticed that we will not need to invoke log_flush_task.wait() for the LRU eviction flushing. buf_flush_list_space(): Replaces buf_flush_dirty_pages(). This is like buf_flush_list(), but operating on a single tablespace at a time. Writes at most innodb_io_capacity pages. Returns whether some of the tablespace might remain in the buffer pool.
2021-06-23 11:14:26 +02:00
pthread_cond_broadcast(&buf_pool.done_flush_LRU);
MDEV-26004 Excessive wait times in buf_LRU_get_free_block() buf_LRU_get_free_block(): Initially wait for a single block to be freed, signaled by buf_pool.done_free. Only if that fails and no LRU eviction flushing batch is already running, we initiate a flushing batch that should serve all threads that are currently waiting in buf_LRU_get_free_block(). Note: In an extreme case, this may introduce a performance regression at larger numbers of connections. We observed this in sysbench oltp_update_index with 512MiB buffer pool, 4GiB of data on fast NVMe, and 1000 concurrent connections, on a 20-thread CPU. The contention point appears to be buf_pool.mutex, and the improvement would turn into a regression somewhere beyond 32 concurrent connections. On slower storage, such regression was not observed; instead, the throughput was improving and maximum latency was reduced. The excessive waits were pointed out by Vladislav Vaintroub.
2021-06-24 10:01:18 +02:00
pthread_cond_signal(&buf_pool.done_free);
}
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
buf_dblwr.flush_buffered_writes();
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-25801: buf_flush_dirty_pages() is very slow In commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) the implementation of buf_flush_dirty_pages() was replaced with a slow one, which would perform excessive scans of the buf_pool.flush_list and make little progress. buf_flush_list(), buf_flush_LRU(): Split from buf_flush_lists(). Vladislav Vaintroub noticed that we will not need to invoke log_flush_task.wait() for the LRU eviction flushing. buf_flush_list_space(): Replaces buf_flush_dirty_pages(). This is like buf_flush_list(), but operating on a single tablespace at a time. Writes at most innodb_io_capacity pages. Returns whether some of the tablespace might remain in the buffer pool.
2021-06-23 11:14:26 +02:00
DBUG_PRINT("ib_buf", ("LRU flush completed, " ULINTPF " pages", n_flushed));
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
return n_flushed;
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
}
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
/** Initiate a log checkpoint, discarding the start of the log.
@param oldest_lsn the checkpoint LSN
@param end_lsn log_sys.get_lsn()
@return true if success, false if a checkpoint write was already running */
static bool log_checkpoint_low(lsn_t oldest_lsn, lsn_t end_lsn)
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
{
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
ut_ad(!srv_read_only_mode);
MDEV-23855: Use normal mutex for log_sys.mutex, log_sys.flush_order_mutex With an unreasonably small innodb_log_file_size, the page cleaner thread would frequently acquire log_sys.flush_order_mutex and spend a significant portion of CPU time spinning on that mutex when determining the checkpoint LSN.
2020-10-26 15:43:52 +01:00
mysql_mutex_assert_owner(&log_sys.mutex);
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
ut_ad(oldest_lsn <= end_lsn);
ut_ad(end_lsn == log_sys.get_lsn());
ut_ad(oldest_lsn >= log_sys.last_checkpoint_lsn);
MDEV-28185 InnoDB generates redundant log checkpoints The comparison on the checkpoint age (number of log bytes written since the previous checkpoint) is inaccurate, because the previous FILE_CHECKPOINT record could span two 512-byte log blocks, which will cause the LSN to increase by the size of the log block header and footer. We will still generate a redudant checkpoint if the previous checkpoint wrote some FILE_MODIFY records before the FILE_CHECKPOINT record.
2022-03-29 18:42:10 +02:00
const lsn_t age= oldest_lsn - log_sys.last_checkpoint_lsn;
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
MDEV-28185 InnoDB generates redundant log checkpoints The comparison on the checkpoint age (number of log bytes written since the previous checkpoint) is inaccurate, because the previous FILE_CHECKPOINT record could span two 512-byte log blocks, which will cause the LSN to increase by the size of the log block header and footer. We will still generate a redudant checkpoint if the previous checkpoint wrote some FILE_MODIFY records before the FILE_CHECKPOINT record.
2022-03-29 18:42:10 +02:00
if (age > SIZE_OF_FILE_CHECKPOINT + log_sys.framing_size())
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
/* Some log has been written since the previous checkpoint. */;
MDEV-28185 InnoDB generates redundant log checkpoints The comparison on the checkpoint age (number of log bytes written since the previous checkpoint) is inaccurate, because the previous FILE_CHECKPOINT record could span two 512-byte log blocks, which will cause the LSN to increase by the size of the log block header and footer. We will still generate a redudant checkpoint if the previous checkpoint wrote some FILE_MODIFY records before the FILE_CHECKPOINT record.
2022-03-29 18:42:10 +02:00
else if (age > SIZE_OF_FILE_CHECKPOINT &&
!((log_sys.log.calc_lsn_offset(oldest_lsn) ^
log_sys.log.calc_lsn_offset(log_sys.last_checkpoint_lsn)) &
~lsn_t{OS_FILE_LOG_BLOCK_SIZE - 1}))
/* Some log has been written to the same log block. */;
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
else if (srv_shutdown_state > SRV_SHUTDOWN_INITIATED)
/* MariaDB startup expects the redo log file to be logically empty
(not even containing a FILE_CHECKPOINT record) after a clean shutdown.
Perform an extra checkpoint at shutdown. */;
else
{
/* Do nothing, because nothing was logged (other than a
FILE_CHECKPOINT record) since the previous checkpoint. */
MDEV-23855: Use normal mutex for log_sys.mutex, log_sys.flush_order_mutex With an unreasonably small innodb_log_file_size, the page cleaner thread would frequently acquire log_sys.flush_order_mutex and spend a significant portion of CPU time spinning on that mutex when determining the checkpoint LSN.
2020-10-26 15:43:52 +01:00
mysql_mutex_unlock(&log_sys.mutex);
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
return true;
}
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
MDEV-25975: Merge 10.4 into 10.5
2022-04-06 09:30:49 +02:00
ut_ad(!recv_no_log_write);
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
/* Repeat the FILE_MODIFY records after the checkpoint, in case some
log records between the checkpoint and log_sys.lsn need them.
Finally, write a FILE_CHECKPOINT record. Redo log apply expects to
see a FILE_CHECKPOINT after the checkpoint, except on clean
shutdown, where the log will be empty after the checkpoint.
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
It is important that we write out the redo log before any further
dirty pages are flushed to the tablespace files. At this point,
because we hold log_sys.mutex, mtr_t::commit() in other threads will
be blocked, and no pages can be added to the flush lists. */
lsn_t flush_lsn= oldest_lsn;
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
if (fil_names_clear(flush_lsn, oldest_lsn != end_lsn ||
srv_shutdown_state <= SRV_SHUTDOWN_INITIATED))
{
flush_lsn= log_sys.get_lsn();
ut_ad(flush_lsn >= end_lsn + SIZE_OF_FILE_CHECKPOINT);
MDEV-23855: Use normal mutex for log_sys.mutex, log_sys.flush_order_mutex With an unreasonably small innodb_log_file_size, the page cleaner thread would frequently acquire log_sys.flush_order_mutex and spend a significant portion of CPU time spinning on that mutex when determining the checkpoint LSN.
2020-10-26 15:43:52 +01:00
mysql_mutex_unlock(&log_sys.mutex);
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
log_write_up_to(flush_lsn, true, true);
MDEV-23855: Use normal mutex for log_sys.mutex, log_sys.flush_order_mutex With an unreasonably small innodb_log_file_size, the page cleaner thread would frequently acquire log_sys.flush_order_mutex and spend a significant portion of CPU time spinning on that mutex when determining the checkpoint LSN.
2020-10-26 15:43:52 +01:00
mysql_mutex_lock(&log_sys.mutex);
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
if (log_sys.last_checkpoint_lsn >= oldest_lsn)
{
MDEV-23855: Use normal mutex for log_sys.mutex, log_sys.flush_order_mutex With an unreasonably small innodb_log_file_size, the page cleaner thread would frequently acquire log_sys.flush_order_mutex and spend a significant portion of CPU time spinning on that mutex when determining the checkpoint LSN.
2020-10-26 15:43:52 +01:00
mysql_mutex_unlock(&log_sys.mutex);
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
return true;
}
}
else
ut_ad(oldest_lsn >= log_sys.last_checkpoint_lsn);
ut_ad(log_sys.get_flushed_lsn() >= flush_lsn);
Cleanup: Replace log_sys.n_pending_checkpoint_writes with a Boolean Only one checkpoint may be in progress at a time. The counter log_sys.n_pending_checkpoint_writes was being protected by log_sys.mutex. Let us replace it with the Boolean log_sys.checkpoint_pending.
2022-03-29 13:56:44 +02:00
if (log_sys.checkpoint_pending)
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
{
/* A checkpoint write is running */
MDEV-23855: Use normal mutex for log_sys.mutex, log_sys.flush_order_mutex With an unreasonably small innodb_log_file_size, the page cleaner thread would frequently acquire log_sys.flush_order_mutex and spend a significant portion of CPU time spinning on that mutex when determining the checkpoint LSN.
2020-10-26 15:43:52 +01:00
mysql_mutex_unlock(&log_sys.mutex);
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
return false;
}
log_sys.next_checkpoint_lsn= oldest_lsn;
log_write_checkpoint_info(end_lsn);
MDEV-23855: Use normal mutex for log_sys.mutex, log_sys.flush_order_mutex With an unreasonably small innodb_log_file_size, the page cleaner thread would frequently acquire log_sys.flush_order_mutex and spend a significant portion of CPU time spinning on that mutex when determining the checkpoint LSN.
2020-10-26 15:43:52 +01:00
mysql_mutex_assert_not_owner(&log_sys.mutex);
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
return true;
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
}
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
/** Make a checkpoint. Note that this function does not flush dirty
blocks from the buffer pool: it only checks what is lsn of the oldest
modification in the pool, and writes information about the lsn in
log file. Use log_make_checkpoint() to flush also the pool.
@retval true if the checkpoint was or had been made
@retval false if a checkpoint write was already running */
static bool log_checkpoint()
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
{
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
if (recv_recovery_is_on())
recv_sys.apply(true);
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
switch (srv_file_flush_method) {
case SRV_NOSYNC:
case SRV_O_DIRECT_NO_FSYNC:
break;
default:
fil_flush_file_spaces();
}
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
MDEV-23855: Use normal mutex for log_sys.mutex, log_sys.flush_order_mutex With an unreasonably small innodb_log_file_size, the page cleaner thread would frequently acquire log_sys.flush_order_mutex and spend a significant portion of CPU time spinning on that mutex when determining the checkpoint LSN.
2020-10-26 15:43:52 +01:00
mysql_mutex_lock(&log_sys.mutex);
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
const lsn_t end_lsn= log_sys.get_lsn();
MDEV-23855: Use normal mutex for log_sys.mutex, log_sys.flush_order_mutex With an unreasonably small innodb_log_file_size, the page cleaner thread would frequently acquire log_sys.flush_order_mutex and spend a significant portion of CPU time spinning on that mutex when determining the checkpoint LSN.
2020-10-26 15:43:52 +01:00
mysql_mutex_lock(&log_sys.flush_order_mutex);
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
mysql_mutex_lock(&buf_pool.flush_list_mutex);
const lsn_t oldest_lsn= buf_pool.get_oldest_modification(end_lsn);
mysql_mutex_unlock(&buf_pool.flush_list_mutex);
MDEV-23855: Use normal mutex for log_sys.mutex, log_sys.flush_order_mutex With an unreasonably small innodb_log_file_size, the page cleaner thread would frequently acquire log_sys.flush_order_mutex and spend a significant portion of CPU time spinning on that mutex when determining the checkpoint LSN.
2020-10-26 15:43:52 +01:00
mysql_mutex_unlock(&log_sys.flush_order_mutex);
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
return log_checkpoint_low(oldest_lsn, end_lsn);
}
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
/** Make a checkpoint. */
ATTRIBUTE_COLD void log_make_checkpoint()
{
MDEV-24302 follow-up: RESET MASTER hangs As pointed out by Andrei Elkin, the previous fix did not fix one race condition that may have caused the observed hang. innodb_log_flush_request(): If we are enqueueing the very first request at the same time the log write is being completed, we must ensure that a near-concurrent call to log_flush_notify() will not result in a missed notification. We guarantee this by release-acquire operations on log_requests.start and log_sys.flushed_to_disk_lsn. log_flush_notify_and_unlock(): Cleanup: Always release the mutex. log_sys_t::get_flushed_lsn(): Use acquire memory order. log_sys_t::set_flushed_lsn(): Use release memory order. log_sys_t::set_lsn(): Use release memory order. log_sys_t::get_lsn(): Use relaxed memory order by default, and allow the caller to specify acquire memory order explicitly. Whenever the log_sys.mutex is being held or when log writes are prohibited during startup, we can use a relaxed load. Likewise, in some assertions where reading a stale value of log_sys.lsn should not matter, we can use a relaxed load. This will cause some additional instructions to be emitted on architectures that do not implement Total Store Ordering (TSO), such as POWER, ARM, and RISC-V Weak Memory Ordering (RVWMO).
2021-03-30 08:58:24 +02:00
buf_flush_wait_flushed(log_sys.get_lsn(std::memory_order_acquire));
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
while (!log_checkpoint());
}
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
MDEV-27416 InnoDB hang in buf_flush_wait_flushed(), on log checkpoint InnoDB could sometimes hang when triggering a log checkpoint. This is due to commit 7b1252c03d7131754d9503560fe507b33ca1f8b4 (MDEV-24278), which introduced an untimed wait to buf_flush_page_cleaner(). The hang was noticed by occasional failures of IMPORT TABLESPACE tests, such as innodb.innodb-wl5522, which would (unnecessarily) invoke log_make_checkpoint() from row_import_cleanup(). The reason of the hang was that buf_flush_page_cleaner() would enter untimed sleep despite buf_flush_sync_lsn being set. The exact failure scenario is unclear, because buf_flush_sync_lsn should actually be protected by buf_pool.flush_list_mutex. We prevent the hang by invoking buf_pool.page_cleaner_set_idle(false) whenever we are setting buf_flush_sync_lsn and signaling buf_pool.do_flush_list. The bulk of these changes was originally developed as a preparation for MDEV-26827, to invoke buf_flush_list() from fewer threads, and tested on 10.6 by Matthias Leich. This fix was tested by running 100 repetitions of 100 concurrent instances of the test innodb.innodb-wl5522 on a RelWithDebInfo build, using ext4fs and innodb_flush_method=O_DIRECT on a SATA SSD with 4096-byte block size. During the test, the call to log_make_checkpoint() in row_import_cleanup() was present. buf_flush_list(): Make static. buf_flush_wait(): Wait for buf_pool.get_oldest_modification() to reach a target, by work done in the buf_flush_page_cleaner. If buf_flush_sync_lsn is going to be set, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_ahead(): If buf_flush_sync_lsn or buf_flush_async_lsn is going to be set and the page cleaner woken up, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_wait_flushed(): Invoke buf_flush_wait(). buf_flush_sync(): Invoke recv_sys.apply() at the start in case crash recovery is active. Invoke buf_flush_wait(). buf_flush_sync_batch(): A lower-level variant of buf_flush_sync() that is only called by recv_sys_t::apply(). buf_flush_sync_for_checkpoint(): Do not trigger log apply or checkpoint during recovery. buf_dblwr_t::create(): Only initiate a buffer pool flush, not a checkpoint. row_import_cleanup(): Do not unnecessarily invoke log_make_checkpoint(). Invoking buf_flush_list_space() before starting to generate redo log for the imported tablespace should suffice. srv_prepare_to_delete_redo_log_file(): Set recv_sys.recovery_on in order to prevent buf_flush_sync_for_checkpoint() from initiating a checkpoint while the log is inaccessible. Remove a wait loop that is already part of buf_flush_sync(). Do not invoke fil_names_clear() if the log is being upgraded, because the FILE_MODIFY record is specific to the latest format. create_log_file(): Clear recv_sys.recovery_on only after calling log_make_checkpoint(), to prevent buf_flush_page_cleaner from invoking a checkpoint. innodb_shutdown(): Simplify the logic in mariadb-backup --prepare. os_aio_wait_until_no_pending_writes(): Update the function comment. Apart from row_quiesce_table_start() during FLUSH TABLES...FOR EXPORT, this is being called by buf_flush_list_space(), which is invoked by ALTER TABLE...IMPORT TABLESPACE as well as some encryption operations.
2022-01-04 06:40:31 +01:00
/** Wait for all dirty pages up to an LSN to be written out.
NOTE: The calling thread is not allowed to hold any buffer page latches! */
static void buf_flush_wait(lsn_t lsn)
{
ut_ad(lsn <= log_sys.get_lsn());
while (buf_pool.get_oldest_modification(lsn) < lsn)
{
if (buf_flush_sync_lsn < lsn)
{
buf_flush_sync_lsn= lsn;
buf_pool.page_cleaner_set_idle(false);
pthread_cond_signal(&buf_pool.do_flush_list);
}
my_cond_wait(&buf_pool.done_flush_list,
&buf_pool.flush_list_mutex.m_mutex);
}
}
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
/** Wait until all persistent pages are flushed up to a limit.
@param sync_lsn buf_pool.get_oldest_modification(LSN_MAX) to wait for */
ATTRIBUTE_COLD void buf_flush_wait_flushed(lsn_t sync_lsn)
{
ut_ad(sync_lsn);
ut_ad(sync_lsn < LSN_MAX);
MDEV-23855: Use normal mutex for log_sys.mutex, log_sys.flush_order_mutex With an unreasonably small innodb_log_file_size, the page cleaner thread would frequently acquire log_sys.flush_order_mutex and spend a significant portion of CPU time spinning on that mutex when determining the checkpoint LSN.
2020-10-26 15:43:52 +01:00
mysql_mutex_assert_not_owner(&log_sys.mutex);
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
ut_ad(!srv_read_only_mode);
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
if (recv_recovery_is_on())
recv_sys.apply(true);
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
mysql_mutex_lock(&buf_pool.flush_list_mutex);
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
MDEV-24109 InnoDB hangs with innodb_flush_sync=OFF MDEV-23855 broke the handling of innodb_flush_sync=OFF. That parameter is supposed to limit the page write rate in case the log capacity is being exceeded and log checkpoints are needed. With this fix, the following should pass: ./mtr --mysqld=--loose-innodb-flush-sync=0 One of our best regression tests for page flushing is encryption.innochecksum. With innodb_page_size=16k and innodb_flush_sync=OFF it would likely hang without this fix. log_sys.last_checkpoint_lsn: Declare as Atomic_relaxed<lsn_t> so that we are allowed to read the value while not holding log_sys.mutex. buf_flush_wait_flushed(): Let the page cleaner perform the flushing also if innodb_flush_sync=OFF. After the page cleaner has completed, perform a checkpoint if it is needed, because buf_flush_sync_for_checkpoint() will not be run if innodb_flush_sync=OFF. buf_flush_ahead(): Simplify the condition. We do not really care whether buf_flush_page_cleaner() is running. buf_flush_page_cleaner(): Evaluate innodb_flush_sync at the low level. If innodb_flush_sync=OFF, rate-limit the batches to innodb_io_capacity_max pages per second. Reviewed by: Vladislav Vaintroub
2020-11-04 15:55:36 +01:00
if (buf_pool.get_oldest_modification(sync_lsn) < sync_lsn)
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
{
MDEV-27416 InnoDB hang in buf_flush_wait_flushed(), on log checkpoint InnoDB could sometimes hang when triggering a log checkpoint. This is due to commit 7b1252c03d7131754d9503560fe507b33ca1f8b4 (MDEV-24278), which introduced an untimed wait to buf_flush_page_cleaner(). The hang was noticed by occasional failures of IMPORT TABLESPACE tests, such as innodb.innodb-wl5522, which would (unnecessarily) invoke log_make_checkpoint() from row_import_cleanup(). The reason of the hang was that buf_flush_page_cleaner() would enter untimed sleep despite buf_flush_sync_lsn being set. The exact failure scenario is unclear, because buf_flush_sync_lsn should actually be protected by buf_pool.flush_list_mutex. We prevent the hang by invoking buf_pool.page_cleaner_set_idle(false) whenever we are setting buf_flush_sync_lsn and signaling buf_pool.do_flush_list. The bulk of these changes was originally developed as a preparation for MDEV-26827, to invoke buf_flush_list() from fewer threads, and tested on 10.6 by Matthias Leich. This fix was tested by running 100 repetitions of 100 concurrent instances of the test innodb.innodb-wl5522 on a RelWithDebInfo build, using ext4fs and innodb_flush_method=O_DIRECT on a SATA SSD with 4096-byte block size. During the test, the call to log_make_checkpoint() in row_import_cleanup() was present. buf_flush_list(): Make static. buf_flush_wait(): Wait for buf_pool.get_oldest_modification() to reach a target, by work done in the buf_flush_page_cleaner. If buf_flush_sync_lsn is going to be set, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_ahead(): If buf_flush_sync_lsn or buf_flush_async_lsn is going to be set and the page cleaner woken up, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_wait_flushed(): Invoke buf_flush_wait(). buf_flush_sync(): Invoke recv_sys.apply() at the start in case crash recovery is active. Invoke buf_flush_wait(). buf_flush_sync_batch(): A lower-level variant of buf_flush_sync() that is only called by recv_sys_t::apply(). buf_flush_sync_for_checkpoint(): Do not trigger log apply or checkpoint during recovery. buf_dblwr_t::create(): Only initiate a buffer pool flush, not a checkpoint. row_import_cleanup(): Do not unnecessarily invoke log_make_checkpoint(). Invoking buf_flush_list_space() before starting to generate redo log for the imported tablespace should suffice. srv_prepare_to_delete_redo_log_file(): Set recv_sys.recovery_on in order to prevent buf_flush_sync_for_checkpoint() from initiating a checkpoint while the log is inaccessible. Remove a wait loop that is already part of buf_flush_sync(). Do not invoke fil_names_clear() if the log is being upgraded, because the FILE_MODIFY record is specific to the latest format. create_log_file(): Clear recv_sys.recovery_on only after calling log_make_checkpoint(), to prevent buf_flush_page_cleaner from invoking a checkpoint. innodb_shutdown(): Simplify the logic in mariadb-backup --prepare. os_aio_wait_until_no_pending_writes(): Update the function comment. Apart from row_quiesce_table_start() during FLUSH TABLES...FOR EXPORT, this is being called by buf_flush_list_space(), which is invoked by ALTER TABLE...IMPORT TABLESPACE as well as some encryption operations.
2022-01-04 06:40:31 +01:00
MONITOR_INC(MONITOR_FLUSH_SYNC_WAITS);
MDEV-24109 InnoDB hangs with innodb_flush_sync=OFF MDEV-23855 broke the handling of innodb_flush_sync=OFF. That parameter is supposed to limit the page write rate in case the log capacity is being exceeded and log checkpoints are needed. With this fix, the following should pass: ./mtr --mysqld=--loose-innodb-flush-sync=0 One of our best regression tests for page flushing is encryption.innochecksum. With innodb_page_size=16k and innodb_flush_sync=OFF it would likely hang without this fix. log_sys.last_checkpoint_lsn: Declare as Atomic_relaxed<lsn_t> so that we are allowed to read the value while not holding log_sys.mutex. buf_flush_wait_flushed(): Let the page cleaner perform the flushing also if innodb_flush_sync=OFF. After the page cleaner has completed, perform a checkpoint if it is needed, because buf_flush_sync_for_checkpoint() will not be run if innodb_flush_sync=OFF. buf_flush_ahead(): Simplify the condition. We do not really care whether buf_flush_page_cleaner() is running. buf_flush_page_cleaner(): Evaluate innodb_flush_sync at the low level. If innodb_flush_sync=OFF, rate-limit the batches to innodb_io_capacity_max pages per second. Reviewed by: Vladislav Vaintroub
2020-11-04 15:55:36 +01:00
#if 1 /* FIXME: remove this, and guarantee that the page cleaner serves us */
MDEV-27416 InnoDB hang in buf_flush_wait_flushed(), on log checkpoint InnoDB could sometimes hang when triggering a log checkpoint. This is due to commit 7b1252c03d7131754d9503560fe507b33ca1f8b4 (MDEV-24278), which introduced an untimed wait to buf_flush_page_cleaner(). The hang was noticed by occasional failures of IMPORT TABLESPACE tests, such as innodb.innodb-wl5522, which would (unnecessarily) invoke log_make_checkpoint() from row_import_cleanup(). The reason of the hang was that buf_flush_page_cleaner() would enter untimed sleep despite buf_flush_sync_lsn being set. The exact failure scenario is unclear, because buf_flush_sync_lsn should actually be protected by buf_pool.flush_list_mutex. We prevent the hang by invoking buf_pool.page_cleaner_set_idle(false) whenever we are setting buf_flush_sync_lsn and signaling buf_pool.do_flush_list. The bulk of these changes was originally developed as a preparation for MDEV-26827, to invoke buf_flush_list() from fewer threads, and tested on 10.6 by Matthias Leich. This fix was tested by running 100 repetitions of 100 concurrent instances of the test innodb.innodb-wl5522 on a RelWithDebInfo build, using ext4fs and innodb_flush_method=O_DIRECT on a SATA SSD with 4096-byte block size. During the test, the call to log_make_checkpoint() in row_import_cleanup() was present. buf_flush_list(): Make static. buf_flush_wait(): Wait for buf_pool.get_oldest_modification() to reach a target, by work done in the buf_flush_page_cleaner. If buf_flush_sync_lsn is going to be set, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_ahead(): If buf_flush_sync_lsn or buf_flush_async_lsn is going to be set and the page cleaner woken up, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_wait_flushed(): Invoke buf_flush_wait(). buf_flush_sync(): Invoke recv_sys.apply() at the start in case crash recovery is active. Invoke buf_flush_wait(). buf_flush_sync_batch(): A lower-level variant of buf_flush_sync() that is only called by recv_sys_t::apply(). buf_flush_sync_for_checkpoint(): Do not trigger log apply or checkpoint during recovery. buf_dblwr_t::create(): Only initiate a buffer pool flush, not a checkpoint. row_import_cleanup(): Do not unnecessarily invoke log_make_checkpoint(). Invoking buf_flush_list_space() before starting to generate redo log for the imported tablespace should suffice. srv_prepare_to_delete_redo_log_file(): Set recv_sys.recovery_on in order to prevent buf_flush_sync_for_checkpoint() from initiating a checkpoint while the log is inaccessible. Remove a wait loop that is already part of buf_flush_sync(). Do not invoke fil_names_clear() if the log is being upgraded, because the FILE_MODIFY record is specific to the latest format. create_log_file(): Clear recv_sys.recovery_on only after calling log_make_checkpoint(), to prevent buf_flush_page_cleaner from invoking a checkpoint. innodb_shutdown(): Simplify the logic in mariadb-backup --prepare. os_aio_wait_until_no_pending_writes(): Update the function comment. Apart from row_quiesce_table_start() during FLUSH TABLES...FOR EXPORT, this is being called by buf_flush_list_space(), which is invoked by ALTER TABLE...IMPORT TABLESPACE as well as some encryption operations.
2022-01-04 06:40:31 +01:00
if (UNIV_UNLIKELY(!buf_page_cleaner_is_active))
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
{
MDEV-24109 InnoDB hangs with innodb_flush_sync=OFF MDEV-23855 broke the handling of innodb_flush_sync=OFF. That parameter is supposed to limit the page write rate in case the log capacity is being exceeded and log checkpoints are needed. With this fix, the following should pass: ./mtr --mysqld=--loose-innodb-flush-sync=0 One of our best regression tests for page flushing is encryption.innochecksum. With innodb_page_size=16k and innodb_flush_sync=OFF it would likely hang without this fix. log_sys.last_checkpoint_lsn: Declare as Atomic_relaxed<lsn_t> so that we are allowed to read the value while not holding log_sys.mutex. buf_flush_wait_flushed(): Let the page cleaner perform the flushing also if innodb_flush_sync=OFF. After the page cleaner has completed, perform a checkpoint if it is needed, because buf_flush_sync_for_checkpoint() will not be run if innodb_flush_sync=OFF. buf_flush_ahead(): Simplify the condition. We do not really care whether buf_flush_page_cleaner() is running. buf_flush_page_cleaner(): Evaluate innodb_flush_sync at the low level. If innodb_flush_sync=OFF, rate-limit the batches to innodb_io_capacity_max pages per second. Reviewed by: Vladislav Vaintroub
2020-11-04 15:55:36 +01:00
do
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
{
MDEV-24109 InnoDB hangs with innodb_flush_sync=OFF MDEV-23855 broke the handling of innodb_flush_sync=OFF. That parameter is supposed to limit the page write rate in case the log capacity is being exceeded and log checkpoints are needed. With this fix, the following should pass: ./mtr --mysqld=--loose-innodb-flush-sync=0 One of our best regression tests for page flushing is encryption.innochecksum. With innodb_page_size=16k and innodb_flush_sync=OFF it would likely hang without this fix. log_sys.last_checkpoint_lsn: Declare as Atomic_relaxed<lsn_t> so that we are allowed to read the value while not holding log_sys.mutex. buf_flush_wait_flushed(): Let the page cleaner perform the flushing also if innodb_flush_sync=OFF. After the page cleaner has completed, perform a checkpoint if it is needed, because buf_flush_sync_for_checkpoint() will not be run if innodb_flush_sync=OFF. buf_flush_ahead(): Simplify the condition. We do not really care whether buf_flush_page_cleaner() is running. buf_flush_page_cleaner(): Evaluate innodb_flush_sync at the low level. If innodb_flush_sync=OFF, rate-limit the batches to innodb_io_capacity_max pages per second. Reviewed by: Vladislav Vaintroub
2020-11-04 15:55:36 +01:00
mysql_mutex_unlock(&buf_pool.flush_list_mutex);
MDEV-25801: buf_flush_dirty_pages() is very slow In commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) the implementation of buf_flush_dirty_pages() was replaced with a slow one, which would perform excessive scans of the buf_pool.flush_list and make little progress. buf_flush_list(), buf_flush_LRU(): Split from buf_flush_lists(). Vladislav Vaintroub noticed that we will not need to invoke log_flush_task.wait() for the LRU eviction flushing. buf_flush_list_space(): Replaces buf_flush_dirty_pages(). This is like buf_flush_list(), but operating on a single tablespace at a time. Writes at most innodb_io_capacity pages. Returns whether some of the tablespace might remain in the buffer pool.
2021-06-23 11:14:26 +02:00
ulint n_pages= buf_flush_list(srv_max_io_capacity, sync_lsn);
MDEV-24109 InnoDB hangs with innodb_flush_sync=OFF MDEV-23855 broke the handling of innodb_flush_sync=OFF. That parameter is supposed to limit the page write rate in case the log capacity is being exceeded and log checkpoints are needed. With this fix, the following should pass: ./mtr --mysqld=--loose-innodb-flush-sync=0 One of our best regression tests for page flushing is encryption.innochecksum. With innodb_page_size=16k and innodb_flush_sync=OFF it would likely hang without this fix. log_sys.last_checkpoint_lsn: Declare as Atomic_relaxed<lsn_t> so that we are allowed to read the value while not holding log_sys.mutex. buf_flush_wait_flushed(): Let the page cleaner perform the flushing also if innodb_flush_sync=OFF. After the page cleaner has completed, perform a checkpoint if it is needed, because buf_flush_sync_for_checkpoint() will not be run if innodb_flush_sync=OFF. buf_flush_ahead(): Simplify the condition. We do not really care whether buf_flush_page_cleaner() is running. buf_flush_page_cleaner(): Evaluate innodb_flush_sync at the low level. If innodb_flush_sync=OFF, rate-limit the batches to innodb_io_capacity_max pages per second. Reviewed by: Vladislav Vaintroub
2020-11-04 15:55:36 +01:00
buf_flush_wait_batch_end_acquiring_mutex(false);
if (n_pages)
{
MONITOR_INC_VALUE_CUMULATIVE(MONITOR_FLUSH_SYNC_TOTAL_PAGE,
MONITOR_FLUSH_SYNC_COUNT,
MONITOR_FLUSH_SYNC_PAGES, n_pages);
}
mysql_mutex_lock(&buf_pool.flush_list_mutex);
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
}
MDEV-24109 InnoDB hangs with innodb_flush_sync=OFF MDEV-23855 broke the handling of innodb_flush_sync=OFF. That parameter is supposed to limit the page write rate in case the log capacity is being exceeded and log checkpoints are needed. With this fix, the following should pass: ./mtr --mysqld=--loose-innodb-flush-sync=0 One of our best regression tests for page flushing is encryption.innochecksum. With innodb_page_size=16k and innodb_flush_sync=OFF it would likely hang without this fix. log_sys.last_checkpoint_lsn: Declare as Atomic_relaxed<lsn_t> so that we are allowed to read the value while not holding log_sys.mutex. buf_flush_wait_flushed(): Let the page cleaner perform the flushing also if innodb_flush_sync=OFF. After the page cleaner has completed, perform a checkpoint if it is needed, because buf_flush_sync_for_checkpoint() will not be run if innodb_flush_sync=OFF. buf_flush_ahead(): Simplify the condition. We do not really care whether buf_flush_page_cleaner() is running. buf_flush_page_cleaner(): Evaluate innodb_flush_sync at the low level. If innodb_flush_sync=OFF, rate-limit the batches to innodb_io_capacity_max pages per second. Reviewed by: Vladislav Vaintroub
2020-11-04 15:55:36 +01:00
while (buf_pool.get_oldest_modification(sync_lsn) < sync_lsn);
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
}
MDEV-27416 InnoDB hang in buf_flush_wait_flushed(), on log checkpoint InnoDB could sometimes hang when triggering a log checkpoint. This is due to commit 7b1252c03d7131754d9503560fe507b33ca1f8b4 (MDEV-24278), which introduced an untimed wait to buf_flush_page_cleaner(). The hang was noticed by occasional failures of IMPORT TABLESPACE tests, such as innodb.innodb-wl5522, which would (unnecessarily) invoke log_make_checkpoint() from row_import_cleanup(). The reason of the hang was that buf_flush_page_cleaner() would enter untimed sleep despite buf_flush_sync_lsn being set. The exact failure scenario is unclear, because buf_flush_sync_lsn should actually be protected by buf_pool.flush_list_mutex. We prevent the hang by invoking buf_pool.page_cleaner_set_idle(false) whenever we are setting buf_flush_sync_lsn and signaling buf_pool.do_flush_list. The bulk of these changes was originally developed as a preparation for MDEV-26827, to invoke buf_flush_list() from fewer threads, and tested on 10.6 by Matthias Leich. This fix was tested by running 100 repetitions of 100 concurrent instances of the test innodb.innodb-wl5522 on a RelWithDebInfo build, using ext4fs and innodb_flush_method=O_DIRECT on a SATA SSD with 4096-byte block size. During the test, the call to log_make_checkpoint() in row_import_cleanup() was present. buf_flush_list(): Make static. buf_flush_wait(): Wait for buf_pool.get_oldest_modification() to reach a target, by work done in the buf_flush_page_cleaner. If buf_flush_sync_lsn is going to be set, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_ahead(): If buf_flush_sync_lsn or buf_flush_async_lsn is going to be set and the page cleaner woken up, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_wait_flushed(): Invoke buf_flush_wait(). buf_flush_sync(): Invoke recv_sys.apply() at the start in case crash recovery is active. Invoke buf_flush_wait(). buf_flush_sync_batch(): A lower-level variant of buf_flush_sync() that is only called by recv_sys_t::apply(). buf_flush_sync_for_checkpoint(): Do not trigger log apply or checkpoint during recovery. buf_dblwr_t::create(): Only initiate a buffer pool flush, not a checkpoint. row_import_cleanup(): Do not unnecessarily invoke log_make_checkpoint(). Invoking buf_flush_list_space() before starting to generate redo log for the imported tablespace should suffice. srv_prepare_to_delete_redo_log_file(): Set recv_sys.recovery_on in order to prevent buf_flush_sync_for_checkpoint() from initiating a checkpoint while the log is inaccessible. Remove a wait loop that is already part of buf_flush_sync(). Do not invoke fil_names_clear() if the log is being upgraded, because the FILE_MODIFY record is specific to the latest format. create_log_file(): Clear recv_sys.recovery_on only after calling log_make_checkpoint(), to prevent buf_flush_page_cleaner from invoking a checkpoint. innodb_shutdown(): Simplify the logic in mariadb-backup --prepare. os_aio_wait_until_no_pending_writes(): Update the function comment. Apart from row_quiesce_table_start() during FLUSH TABLES...FOR EXPORT, this is being called by buf_flush_list_space(), which is invoked by ALTER TABLE...IMPORT TABLESPACE as well as some encryption operations.
2022-01-04 06:40:31 +01:00
else
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
#endif
MDEV-24109 InnoDB hangs with innodb_flush_sync=OFF MDEV-23855 broke the handling of innodb_flush_sync=OFF. That parameter is supposed to limit the page write rate in case the log capacity is being exceeded and log checkpoints are needed. With this fix, the following should pass: ./mtr --mysqld=--loose-innodb-flush-sync=0 One of our best regression tests for page flushing is encryption.innochecksum. With innodb_page_size=16k and innodb_flush_sync=OFF it would likely hang without this fix. log_sys.last_checkpoint_lsn: Declare as Atomic_relaxed<lsn_t> so that we are allowed to read the value while not holding log_sys.mutex. buf_flush_wait_flushed(): Let the page cleaner perform the flushing also if innodb_flush_sync=OFF. After the page cleaner has completed, perform a checkpoint if it is needed, because buf_flush_sync_for_checkpoint() will not be run if innodb_flush_sync=OFF. buf_flush_ahead(): Simplify the condition. We do not really care whether buf_flush_page_cleaner() is running. buf_flush_page_cleaner(): Evaluate innodb_flush_sync at the low level. If innodb_flush_sync=OFF, rate-limit the batches to innodb_io_capacity_max pages per second. Reviewed by: Vladislav Vaintroub
2020-11-04 15:55:36 +01:00
{
thd_wait_begin(nullptr, THD_WAIT_DISKIO);
MDEV-27416 InnoDB hang in buf_flush_wait_flushed(), on log checkpoint InnoDB could sometimes hang when triggering a log checkpoint. This is due to commit 7b1252c03d7131754d9503560fe507b33ca1f8b4 (MDEV-24278), which introduced an untimed wait to buf_flush_page_cleaner(). The hang was noticed by occasional failures of IMPORT TABLESPACE tests, such as innodb.innodb-wl5522, which would (unnecessarily) invoke log_make_checkpoint() from row_import_cleanup(). The reason of the hang was that buf_flush_page_cleaner() would enter untimed sleep despite buf_flush_sync_lsn being set. The exact failure scenario is unclear, because buf_flush_sync_lsn should actually be protected by buf_pool.flush_list_mutex. We prevent the hang by invoking buf_pool.page_cleaner_set_idle(false) whenever we are setting buf_flush_sync_lsn and signaling buf_pool.do_flush_list. The bulk of these changes was originally developed as a preparation for MDEV-26827, to invoke buf_flush_list() from fewer threads, and tested on 10.6 by Matthias Leich. This fix was tested by running 100 repetitions of 100 concurrent instances of the test innodb.innodb-wl5522 on a RelWithDebInfo build, using ext4fs and innodb_flush_method=O_DIRECT on a SATA SSD with 4096-byte block size. During the test, the call to log_make_checkpoint() in row_import_cleanup() was present. buf_flush_list(): Make static. buf_flush_wait(): Wait for buf_pool.get_oldest_modification() to reach a target, by work done in the buf_flush_page_cleaner. If buf_flush_sync_lsn is going to be set, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_ahead(): If buf_flush_sync_lsn or buf_flush_async_lsn is going to be set and the page cleaner woken up, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_wait_flushed(): Invoke buf_flush_wait(). buf_flush_sync(): Invoke recv_sys.apply() at the start in case crash recovery is active. Invoke buf_flush_wait(). buf_flush_sync_batch(): A lower-level variant of buf_flush_sync() that is only called by recv_sys_t::apply(). buf_flush_sync_for_checkpoint(): Do not trigger log apply or checkpoint during recovery. buf_dblwr_t::create(): Only initiate a buffer pool flush, not a checkpoint. row_import_cleanup(): Do not unnecessarily invoke log_make_checkpoint(). Invoking buf_flush_list_space() before starting to generate redo log for the imported tablespace should suffice. srv_prepare_to_delete_redo_log_file(): Set recv_sys.recovery_on in order to prevent buf_flush_sync_for_checkpoint() from initiating a checkpoint while the log is inaccessible. Remove a wait loop that is already part of buf_flush_sync(). Do not invoke fil_names_clear() if the log is being upgraded, because the FILE_MODIFY record is specific to the latest format. create_log_file(): Clear recv_sys.recovery_on only after calling log_make_checkpoint(), to prevent buf_flush_page_cleaner from invoking a checkpoint. innodb_shutdown(): Simplify the logic in mariadb-backup --prepare. os_aio_wait_until_no_pending_writes(): Update the function comment. Apart from row_quiesce_table_start() during FLUSH TABLES...FOR EXPORT, this is being called by buf_flush_list_space(), which is invoked by ALTER TABLE...IMPORT TABLESPACE as well as some encryption operations.
2022-01-04 06:40:31 +01:00
tpool::tpool_wait_begin();
buf_flush_wait(sync_lsn);
MDEV-24109 InnoDB hangs with innodb_flush_sync=OFF MDEV-23855 broke the handling of innodb_flush_sync=OFF. That parameter is supposed to limit the page write rate in case the log capacity is being exceeded and log checkpoints are needed. With this fix, the following should pass: ./mtr --mysqld=--loose-innodb-flush-sync=0 One of our best regression tests for page flushing is encryption.innochecksum. With innodb_page_size=16k and innodb_flush_sync=OFF it would likely hang without this fix. log_sys.last_checkpoint_lsn: Declare as Atomic_relaxed<lsn_t> so that we are allowed to read the value while not holding log_sys.mutex. buf_flush_wait_flushed(): Let the page cleaner perform the flushing also if innodb_flush_sync=OFF. After the page cleaner has completed, perform a checkpoint if it is needed, because buf_flush_sync_for_checkpoint() will not be run if innodb_flush_sync=OFF. buf_flush_ahead(): Simplify the condition. We do not really care whether buf_flush_page_cleaner() is running. buf_flush_page_cleaner(): Evaluate innodb_flush_sync at the low level. If innodb_flush_sync=OFF, rate-limit the batches to innodb_io_capacity_max pages per second. Reviewed by: Vladislav Vaintroub
2020-11-04 15:55:36 +01:00
tpool::tpool_wait_end();
MDEV-27416 InnoDB hang in buf_flush_wait_flushed(), on log checkpoint InnoDB could sometimes hang when triggering a log checkpoint. This is due to commit 7b1252c03d7131754d9503560fe507b33ca1f8b4 (MDEV-24278), which introduced an untimed wait to buf_flush_page_cleaner(). The hang was noticed by occasional failures of IMPORT TABLESPACE tests, such as innodb.innodb-wl5522, which would (unnecessarily) invoke log_make_checkpoint() from row_import_cleanup(). The reason of the hang was that buf_flush_page_cleaner() would enter untimed sleep despite buf_flush_sync_lsn being set. The exact failure scenario is unclear, because buf_flush_sync_lsn should actually be protected by buf_pool.flush_list_mutex. We prevent the hang by invoking buf_pool.page_cleaner_set_idle(false) whenever we are setting buf_flush_sync_lsn and signaling buf_pool.do_flush_list. The bulk of these changes was originally developed as a preparation for MDEV-26827, to invoke buf_flush_list() from fewer threads, and tested on 10.6 by Matthias Leich. This fix was tested by running 100 repetitions of 100 concurrent instances of the test innodb.innodb-wl5522 on a RelWithDebInfo build, using ext4fs and innodb_flush_method=O_DIRECT on a SATA SSD with 4096-byte block size. During the test, the call to log_make_checkpoint() in row_import_cleanup() was present. buf_flush_list(): Make static. buf_flush_wait(): Wait for buf_pool.get_oldest_modification() to reach a target, by work done in the buf_flush_page_cleaner. If buf_flush_sync_lsn is going to be set, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_ahead(): If buf_flush_sync_lsn or buf_flush_async_lsn is going to be set and the page cleaner woken up, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_wait_flushed(): Invoke buf_flush_wait(). buf_flush_sync(): Invoke recv_sys.apply() at the start in case crash recovery is active. Invoke buf_flush_wait(). buf_flush_sync_batch(): A lower-level variant of buf_flush_sync() that is only called by recv_sys_t::apply(). buf_flush_sync_for_checkpoint(): Do not trigger log apply or checkpoint during recovery. buf_dblwr_t::create(): Only initiate a buffer pool flush, not a checkpoint. row_import_cleanup(): Do not unnecessarily invoke log_make_checkpoint(). Invoking buf_flush_list_space() before starting to generate redo log for the imported tablespace should suffice. srv_prepare_to_delete_redo_log_file(): Set recv_sys.recovery_on in order to prevent buf_flush_sync_for_checkpoint() from initiating a checkpoint while the log is inaccessible. Remove a wait loop that is already part of buf_flush_sync(). Do not invoke fil_names_clear() if the log is being upgraded, because the FILE_MODIFY record is specific to the latest format. create_log_file(): Clear recv_sys.recovery_on only after calling log_make_checkpoint(), to prevent buf_flush_page_cleaner from invoking a checkpoint. innodb_shutdown(): Simplify the logic in mariadb-backup --prepare. os_aio_wait_until_no_pending_writes(): Update the function comment. Apart from row_quiesce_table_start() during FLUSH TABLES...FOR EXPORT, this is being called by buf_flush_list_space(), which is invoked by ALTER TABLE...IMPORT TABLESPACE as well as some encryption operations.
2022-01-04 06:40:31 +01:00
thd_wait_end(nullptr);
MDEV-24109 InnoDB hangs with innodb_flush_sync=OFF MDEV-23855 broke the handling of innodb_flush_sync=OFF. That parameter is supposed to limit the page write rate in case the log capacity is being exceeded and log checkpoints are needed. With this fix, the following should pass: ./mtr --mysqld=--loose-innodb-flush-sync=0 One of our best regression tests for page flushing is encryption.innochecksum. With innodb_page_size=16k and innodb_flush_sync=OFF it would likely hang without this fix. log_sys.last_checkpoint_lsn: Declare as Atomic_relaxed<lsn_t> so that we are allowed to read the value while not holding log_sys.mutex. buf_flush_wait_flushed(): Let the page cleaner perform the flushing also if innodb_flush_sync=OFF. After the page cleaner has completed, perform a checkpoint if it is needed, because buf_flush_sync_for_checkpoint() will not be run if innodb_flush_sync=OFF. buf_flush_ahead(): Simplify the condition. We do not really care whether buf_flush_page_cleaner() is running. buf_flush_page_cleaner(): Evaluate innodb_flush_sync at the low level. If innodb_flush_sync=OFF, rate-limit the batches to innodb_io_capacity_max pages per second. Reviewed by: Vladislav Vaintroub
2020-11-04 15:55:36 +01:00
}
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
}
mysql_mutex_unlock(&buf_pool.flush_list_mutex);
MDEV-24109 InnoDB hangs with innodb_flush_sync=OFF MDEV-23855 broke the handling of innodb_flush_sync=OFF. That parameter is supposed to limit the page write rate in case the log capacity is being exceeded and log checkpoints are needed. With this fix, the following should pass: ./mtr --mysqld=--loose-innodb-flush-sync=0 One of our best regression tests for page flushing is encryption.innochecksum. With innodb_page_size=16k and innodb_flush_sync=OFF it would likely hang without this fix. log_sys.last_checkpoint_lsn: Declare as Atomic_relaxed<lsn_t> so that we are allowed to read the value while not holding log_sys.mutex. buf_flush_wait_flushed(): Let the page cleaner perform the flushing also if innodb_flush_sync=OFF. After the page cleaner has completed, perform a checkpoint if it is needed, because buf_flush_sync_for_checkpoint() will not be run if innodb_flush_sync=OFF. buf_flush_ahead(): Simplify the condition. We do not really care whether buf_flush_page_cleaner() is running. buf_flush_page_cleaner(): Evaluate innodb_flush_sync at the low level. If innodb_flush_sync=OFF, rate-limit the batches to innodb_io_capacity_max pages per second. Reviewed by: Vladislav Vaintroub
2020-11-04 15:55:36 +01:00
if (UNIV_UNLIKELY(log_sys.last_checkpoint_lsn < sync_lsn))
MDEV-25948 Remove log_flush_task Vladislav Vaintroub suggested that invoking log_flush_up_to() for every page could perform better than invoking a log write between buf_pool.flush_list batches, like we started doing in commit 3a9a3be1c64b14c05648e87ebe0f1dd96457de41 (MDEV-23855). This could depend on the sequence in which pages are being modified. The buf_pool.flush_list is ordered by oldest_modification, while the FIL_PAGE_LSN of the pages is theoretically independent of that. In the pathological case, we will wait for a log write before writing each individual page. It turns out that we can defer the call to log_flush_up_to() until just before submitting the page write. If the doublewrite buffer is being used, we can submit a write batch of "future" pages to the doublewrite buffer, and only wait for the log write right before we are writing an already doublewritten page. The next doublewrite batch will not be initiated before the last page write from the current batch has completed. When a future version introduces asynchronous writes if the log, we could initiate a write at the start of a flushing batch, to reduce waiting further.
2021-06-23 11:01:41 +02:00
{
/* If the buffer pool was clean, no log write was guaranteed
to happen until now. There could be an outstanding FILE_CHECKPOINT
record from a previous fil_names_clear() call, which we must
write out before we can advance the checkpoint. */
if (sync_lsn > log_sys.get_flushed_lsn())
log_write_up_to(sync_lsn, true);
MDEV-24109 InnoDB hangs with innodb_flush_sync=OFF MDEV-23855 broke the handling of innodb_flush_sync=OFF. That parameter is supposed to limit the page write rate in case the log capacity is being exceeded and log checkpoints are needed. With this fix, the following should pass: ./mtr --mysqld=--loose-innodb-flush-sync=0 One of our best regression tests for page flushing is encryption.innochecksum. With innodb_page_size=16k and innodb_flush_sync=OFF it would likely hang without this fix. log_sys.last_checkpoint_lsn: Declare as Atomic_relaxed<lsn_t> so that we are allowed to read the value while not holding log_sys.mutex. buf_flush_wait_flushed(): Let the page cleaner perform the flushing also if innodb_flush_sync=OFF. After the page cleaner has completed, perform a checkpoint if it is needed, because buf_flush_sync_for_checkpoint() will not be run if innodb_flush_sync=OFF. buf_flush_ahead(): Simplify the condition. We do not really care whether buf_flush_page_cleaner() is running. buf_flush_page_cleaner(): Evaluate innodb_flush_sync at the low level. If innodb_flush_sync=OFF, rate-limit the batches to innodb_io_capacity_max pages per second. Reviewed by: Vladislav Vaintroub
2020-11-04 15:55:36 +01:00
log_checkpoint();
MDEV-25948 Remove log_flush_task Vladislav Vaintroub suggested that invoking log_flush_up_to() for every page could perform better than invoking a log write between buf_pool.flush_list batches, like we started doing in commit 3a9a3be1c64b14c05648e87ebe0f1dd96457de41 (MDEV-23855). This could depend on the sequence in which pages are being modified. The buf_pool.flush_list is ordered by oldest_modification, while the FIL_PAGE_LSN of the pages is theoretically independent of that. In the pathological case, we will wait for a log write before writing each individual page. It turns out that we can defer the call to log_flush_up_to() until just before submitting the page write. If the doublewrite buffer is being used, we can submit a write batch of "future" pages to the doublewrite buffer, and only wait for the log write right before we are writing an already doublewritten page. The next doublewrite batch will not be initiated before the last page write from the current batch has completed. When a future version introduces asynchronous writes if the log, we could initiate a write at the start of a flushing batch, to reduce waiting further.
2021-06-23 11:01:41 +02:00
}
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
}
MDEV-25113: Introduce a page cleaner mode before 'furious flush' MDEV-23855 changed the way how the page cleaner is signaled by user threads. If a threshold is exceeded, a mini-transaction commit would invoke buf_flush_ahead() in order to initiate page flushing before all writers would eventually grind to halt in log_free_check(), waiting for the checkpoint age to reduce. However, buf_flush_ahead() would always initiate 'furious flushing', making the buf_flush_page_cleaner thread write innodb_io_capacity_max pages per batch, and sleeping no time between batches, until the limit LSN is reached. Because this could saturate the I/O subsystem, system throughput could significantly reduce during these 'furious flushing' spikes. With this change, we introduce a gentler version of flush-ahead, which would write innodb_io_capacity_max pages per second until the 'soft limit' is reached. buf_flush_ahead(): Add a parameter to specify whether furious flushing is requested. buf_flush_async_lsn: Similar to buf_flush_sync_lsn, a limit for the less intrusive flushing. buf_flush_page_cleaner(): Keep working until buf_flush_async_lsn has been reached. log_close(): Suppress a warning message in the event that a new log is being created during startup, when old logs did not exist. Return what type of page cleaning will be needed. mtr_t::finish_write(): Also when m_log.is_small(), invoke log_close(). Return what type of page cleaning will be needed. mtr_t::commit(): Invoke buf_flush_ahead() based on the return value of mtr_t::finish_write().
2021-06-23 12:13:16 +02:00
/** Initiate more eager page flushing if the log checkpoint age is too old.
@param lsn buf_pool.get_oldest_modification(LSN_MAX) target
@param furious true=furious flushing, false=limit to innodb_io_capacity */
ATTRIBUTE_COLD void buf_flush_ahead(lsn_t lsn, bool furious)
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
{
MDEV-23855: Use normal mutex for log_sys.mutex, log_sys.flush_order_mutex With an unreasonably small innodb_log_file_size, the page cleaner thread would frequently acquire log_sys.flush_order_mutex and spend a significant portion of CPU time spinning on that mutex when determining the checkpoint LSN.
2020-10-26 15:43:52 +01:00
mysql_mutex_assert_not_owner(&log_sys.mutex);
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
ut_ad(!srv_read_only_mode);
if (recv_recovery_is_on())
recv_sys.apply(true);
MDEV-25113: Introduce a page cleaner mode before 'furious flush' MDEV-23855 changed the way how the page cleaner is signaled by user threads. If a threshold is exceeded, a mini-transaction commit would invoke buf_flush_ahead() in order to initiate page flushing before all writers would eventually grind to halt in log_free_check(), waiting for the checkpoint age to reduce. However, buf_flush_ahead() would always initiate 'furious flushing', making the buf_flush_page_cleaner thread write innodb_io_capacity_max pages per batch, and sleeping no time between batches, until the limit LSN is reached. Because this could saturate the I/O subsystem, system throughput could significantly reduce during these 'furious flushing' spikes. With this change, we introduce a gentler version of flush-ahead, which would write innodb_io_capacity_max pages per second until the 'soft limit' is reached. buf_flush_ahead(): Add a parameter to specify whether furious flushing is requested. buf_flush_async_lsn: Similar to buf_flush_sync_lsn, a limit for the less intrusive flushing. buf_flush_page_cleaner(): Keep working until buf_flush_async_lsn has been reached. log_close(): Suppress a warning message in the event that a new log is being created during startup, when old logs did not exist. Return what type of page cleaning will be needed. mtr_t::finish_write(): Also when m_log.is_small(), invoke log_close(). Return what type of page cleaning will be needed. mtr_t::commit(): Invoke buf_flush_ahead() based on the return value of mtr_t::finish_write().
2021-06-23 12:13:16 +02:00
Atomic_relaxed<lsn_t> &limit= furious
? buf_flush_sync_lsn : buf_flush_async_lsn;
if (limit < lsn)
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
{
mysql_mutex_lock(&buf_pool.flush_list_mutex);
MDEV-25113: Introduce a page cleaner mode before 'furious flush' MDEV-23855 changed the way how the page cleaner is signaled by user threads. If a threshold is exceeded, a mini-transaction commit would invoke buf_flush_ahead() in order to initiate page flushing before all writers would eventually grind to halt in log_free_check(), waiting for the checkpoint age to reduce. However, buf_flush_ahead() would always initiate 'furious flushing', making the buf_flush_page_cleaner thread write innodb_io_capacity_max pages per batch, and sleeping no time between batches, until the limit LSN is reached. Because this could saturate the I/O subsystem, system throughput could significantly reduce during these 'furious flushing' spikes. With this change, we introduce a gentler version of flush-ahead, which would write innodb_io_capacity_max pages per second until the 'soft limit' is reached. buf_flush_ahead(): Add a parameter to specify whether furious flushing is requested. buf_flush_async_lsn: Similar to buf_flush_sync_lsn, a limit for the less intrusive flushing. buf_flush_page_cleaner(): Keep working until buf_flush_async_lsn has been reached. log_close(): Suppress a warning message in the event that a new log is being created during startup, when old logs did not exist. Return what type of page cleaning will be needed. mtr_t::finish_write(): Also when m_log.is_small(), invoke log_close(). Return what type of page cleaning will be needed. mtr_t::commit(): Invoke buf_flush_ahead() based on the return value of mtr_t::finish_write().
2021-06-23 12:13:16 +02:00
if (limit < lsn)
MDEV-27416 InnoDB hang in buf_flush_wait_flushed(), on log checkpoint InnoDB could sometimes hang when triggering a log checkpoint. This is due to commit 7b1252c03d7131754d9503560fe507b33ca1f8b4 (MDEV-24278), which introduced an untimed wait to buf_flush_page_cleaner(). The hang was noticed by occasional failures of IMPORT TABLESPACE tests, such as innodb.innodb-wl5522, which would (unnecessarily) invoke log_make_checkpoint() from row_import_cleanup(). The reason of the hang was that buf_flush_page_cleaner() would enter untimed sleep despite buf_flush_sync_lsn being set. The exact failure scenario is unclear, because buf_flush_sync_lsn should actually be protected by buf_pool.flush_list_mutex. We prevent the hang by invoking buf_pool.page_cleaner_set_idle(false) whenever we are setting buf_flush_sync_lsn and signaling buf_pool.do_flush_list. The bulk of these changes was originally developed as a preparation for MDEV-26827, to invoke buf_flush_list() from fewer threads, and tested on 10.6 by Matthias Leich. This fix was tested by running 100 repetitions of 100 concurrent instances of the test innodb.innodb-wl5522 on a RelWithDebInfo build, using ext4fs and innodb_flush_method=O_DIRECT on a SATA SSD with 4096-byte block size. During the test, the call to log_make_checkpoint() in row_import_cleanup() was present. buf_flush_list(): Make static. buf_flush_wait(): Wait for buf_pool.get_oldest_modification() to reach a target, by work done in the buf_flush_page_cleaner. If buf_flush_sync_lsn is going to be set, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_ahead(): If buf_flush_sync_lsn or buf_flush_async_lsn is going to be set and the page cleaner woken up, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_wait_flushed(): Invoke buf_flush_wait(). buf_flush_sync(): Invoke recv_sys.apply() at the start in case crash recovery is active. Invoke buf_flush_wait(). buf_flush_sync_batch(): A lower-level variant of buf_flush_sync() that is only called by recv_sys_t::apply(). buf_flush_sync_for_checkpoint(): Do not trigger log apply or checkpoint during recovery. buf_dblwr_t::create(): Only initiate a buffer pool flush, not a checkpoint. row_import_cleanup(): Do not unnecessarily invoke log_make_checkpoint(). Invoking buf_flush_list_space() before starting to generate redo log for the imported tablespace should suffice. srv_prepare_to_delete_redo_log_file(): Set recv_sys.recovery_on in order to prevent buf_flush_sync_for_checkpoint() from initiating a checkpoint while the log is inaccessible. Remove a wait loop that is already part of buf_flush_sync(). Do not invoke fil_names_clear() if the log is being upgraded, because the FILE_MODIFY record is specific to the latest format. create_log_file(): Clear recv_sys.recovery_on only after calling log_make_checkpoint(), to prevent buf_flush_page_cleaner from invoking a checkpoint. innodb_shutdown(): Simplify the logic in mariadb-backup --prepare. os_aio_wait_until_no_pending_writes(): Update the function comment. Apart from row_quiesce_table_start() during FLUSH TABLES...FOR EXPORT, this is being called by buf_flush_list_space(), which is invoked by ALTER TABLE...IMPORT TABLESPACE as well as some encryption operations.
2022-01-04 06:40:31 +01:00
{
MDEV-25113: Introduce a page cleaner mode before 'furious flush' MDEV-23855 changed the way how the page cleaner is signaled by user threads. If a threshold is exceeded, a mini-transaction commit would invoke buf_flush_ahead() in order to initiate page flushing before all writers would eventually grind to halt in log_free_check(), waiting for the checkpoint age to reduce. However, buf_flush_ahead() would always initiate 'furious flushing', making the buf_flush_page_cleaner thread write innodb_io_capacity_max pages per batch, and sleeping no time between batches, until the limit LSN is reached. Because this could saturate the I/O subsystem, system throughput could significantly reduce during these 'furious flushing' spikes. With this change, we introduce a gentler version of flush-ahead, which would write innodb_io_capacity_max pages per second until the 'soft limit' is reached. buf_flush_ahead(): Add a parameter to specify whether furious flushing is requested. buf_flush_async_lsn: Similar to buf_flush_sync_lsn, a limit for the less intrusive flushing. buf_flush_page_cleaner(): Keep working until buf_flush_async_lsn has been reached. log_close(): Suppress a warning message in the event that a new log is being created during startup, when old logs did not exist. Return what type of page cleaning will be needed. mtr_t::finish_write(): Also when m_log.is_small(), invoke log_close(). Return what type of page cleaning will be needed. mtr_t::commit(): Invoke buf_flush_ahead() based on the return value of mtr_t::finish_write().
2021-06-23 12:13:16 +02:00
limit= lsn;
MDEV-27416 InnoDB hang in buf_flush_wait_flushed(), on log checkpoint InnoDB could sometimes hang when triggering a log checkpoint. This is due to commit 7b1252c03d7131754d9503560fe507b33ca1f8b4 (MDEV-24278), which introduced an untimed wait to buf_flush_page_cleaner(). The hang was noticed by occasional failures of IMPORT TABLESPACE tests, such as innodb.innodb-wl5522, which would (unnecessarily) invoke log_make_checkpoint() from row_import_cleanup(). The reason of the hang was that buf_flush_page_cleaner() would enter untimed sleep despite buf_flush_sync_lsn being set. The exact failure scenario is unclear, because buf_flush_sync_lsn should actually be protected by buf_pool.flush_list_mutex. We prevent the hang by invoking buf_pool.page_cleaner_set_idle(false) whenever we are setting buf_flush_sync_lsn and signaling buf_pool.do_flush_list. The bulk of these changes was originally developed as a preparation for MDEV-26827, to invoke buf_flush_list() from fewer threads, and tested on 10.6 by Matthias Leich. This fix was tested by running 100 repetitions of 100 concurrent instances of the test innodb.innodb-wl5522 on a RelWithDebInfo build, using ext4fs and innodb_flush_method=O_DIRECT on a SATA SSD with 4096-byte block size. During the test, the call to log_make_checkpoint() in row_import_cleanup() was present. buf_flush_list(): Make static. buf_flush_wait(): Wait for buf_pool.get_oldest_modification() to reach a target, by work done in the buf_flush_page_cleaner. If buf_flush_sync_lsn is going to be set, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_ahead(): If buf_flush_sync_lsn or buf_flush_async_lsn is going to be set and the page cleaner woken up, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_wait_flushed(): Invoke buf_flush_wait(). buf_flush_sync(): Invoke recv_sys.apply() at the start in case crash recovery is active. Invoke buf_flush_wait(). buf_flush_sync_batch(): A lower-level variant of buf_flush_sync() that is only called by recv_sys_t::apply(). buf_flush_sync_for_checkpoint(): Do not trigger log apply or checkpoint during recovery. buf_dblwr_t::create(): Only initiate a buffer pool flush, not a checkpoint. row_import_cleanup(): Do not unnecessarily invoke log_make_checkpoint(). Invoking buf_flush_list_space() before starting to generate redo log for the imported tablespace should suffice. srv_prepare_to_delete_redo_log_file(): Set recv_sys.recovery_on in order to prevent buf_flush_sync_for_checkpoint() from initiating a checkpoint while the log is inaccessible. Remove a wait loop that is already part of buf_flush_sync(). Do not invoke fil_names_clear() if the log is being upgraded, because the FILE_MODIFY record is specific to the latest format. create_log_file(): Clear recv_sys.recovery_on only after calling log_make_checkpoint(), to prevent buf_flush_page_cleaner from invoking a checkpoint. innodb_shutdown(): Simplify the logic in mariadb-backup --prepare. os_aio_wait_until_no_pending_writes(): Update the function comment. Apart from row_quiesce_table_start() during FLUSH TABLES...FOR EXPORT, this is being called by buf_flush_list_space(), which is invoked by ALTER TABLE...IMPORT TABLESPACE as well as some encryption operations.
2022-01-04 06:40:31 +01:00
buf_pool.page_cleaner_set_idle(false);
pthread_cond_signal(&buf_pool.do_flush_list);
}
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
mysql_mutex_unlock(&buf_pool.flush_list_mutex);
}
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
}
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
/** Wait for pending flushes to complete. */
void buf_flush_wait_batch_end_acquiring_mutex(bool lru)
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
{
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
if (lru ? buf_pool.n_flush_LRU() : buf_pool.n_flush_list())
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
{
mysql_mutex_lock(&buf_pool.mutex);
buf_flush_wait_batch_end(lru);
mysql_mutex_unlock(&buf_pool.mutex);
}
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
}
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
/** Conduct checkpoint-related flushing for innodb_flush_sync=ON,
and try to initiate checkpoints until the target is met.
@param lsn minimum value of buf_pool.get_oldest_modification(LSN_MAX) */
ATTRIBUTE_COLD static void buf_flush_sync_for_checkpoint(lsn_t lsn)
{
ut_ad(!srv_read_only_mode);
for (;;)
{
mysql_mutex_unlock(&buf_pool.flush_list_mutex);
MDEV-25801: buf_flush_dirty_pages() is very slow In commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) the implementation of buf_flush_dirty_pages() was replaced with a slow one, which would perform excessive scans of the buf_pool.flush_list and make little progress. buf_flush_list(), buf_flush_LRU(): Split from buf_flush_lists(). Vladislav Vaintroub noticed that we will not need to invoke log_flush_task.wait() for the LRU eviction flushing. buf_flush_list_space(): Replaces buf_flush_dirty_pages(). This is like buf_flush_list(), but operating on a single tablespace at a time. Writes at most innodb_io_capacity pages. Returns whether some of the tablespace might remain in the buffer pool.
2021-06-23 11:14:26 +02:00
if (ulint n_flushed= buf_flush_list(srv_max_io_capacity, lsn))
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
{
MONITOR_INC_VALUE_CUMULATIVE(MONITOR_FLUSH_SYNC_TOTAL_PAGE,
MONITOR_FLUSH_SYNC_COUNT,
MONITOR_FLUSH_SYNC_PAGES, n_flushed);
}
switch (srv_file_flush_method) {
case SRV_NOSYNC:
case SRV_O_DIRECT_NO_FSYNC:
break;
default:
fil_flush_file_spaces();
}
MDEV-23855: Use normal mutex for log_sys.mutex, log_sys.flush_order_mutex With an unreasonably small innodb_log_file_size, the page cleaner thread would frequently acquire log_sys.flush_order_mutex and spend a significant portion of CPU time spinning on that mutex when determining the checkpoint LSN.
2020-10-26 15:43:52 +01:00
mysql_mutex_lock(&log_sys.mutex);
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
const lsn_t newest_lsn= log_sys.get_lsn();
MDEV-23855: Use normal mutex for log_sys.mutex, log_sys.flush_order_mutex With an unreasonably small innodb_log_file_size, the page cleaner thread would frequently acquire log_sys.flush_order_mutex and spend a significant portion of CPU time spinning on that mutex when determining the checkpoint LSN.
2020-10-26 15:43:52 +01:00
mysql_mutex_lock(&log_sys.flush_order_mutex);
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
mysql_mutex_lock(&buf_pool.flush_list_mutex);
lsn_t measure= buf_pool.get_oldest_modification(0);
MDEV-23855: Use normal mutex for log_sys.mutex, log_sys.flush_order_mutex With an unreasonably small innodb_log_file_size, the page cleaner thread would frequently acquire log_sys.flush_order_mutex and spend a significant portion of CPU time spinning on that mutex when determining the checkpoint LSN.
2020-10-26 15:43:52 +01:00
mysql_mutex_unlock(&log_sys.flush_order_mutex);
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
const lsn_t checkpoint_lsn= measure ? measure : newest_lsn;
MDEV-27416 InnoDB hang in buf_flush_wait_flushed(), on log checkpoint InnoDB could sometimes hang when triggering a log checkpoint. This is due to commit 7b1252c03d7131754d9503560fe507b33ca1f8b4 (MDEV-24278), which introduced an untimed wait to buf_flush_page_cleaner(). The hang was noticed by occasional failures of IMPORT TABLESPACE tests, such as innodb.innodb-wl5522, which would (unnecessarily) invoke log_make_checkpoint() from row_import_cleanup(). The reason of the hang was that buf_flush_page_cleaner() would enter untimed sleep despite buf_flush_sync_lsn being set. The exact failure scenario is unclear, because buf_flush_sync_lsn should actually be protected by buf_pool.flush_list_mutex. We prevent the hang by invoking buf_pool.page_cleaner_set_idle(false) whenever we are setting buf_flush_sync_lsn and signaling buf_pool.do_flush_list. The bulk of these changes was originally developed as a preparation for MDEV-26827, to invoke buf_flush_list() from fewer threads, and tested on 10.6 by Matthias Leich. This fix was tested by running 100 repetitions of 100 concurrent instances of the test innodb.innodb-wl5522 on a RelWithDebInfo build, using ext4fs and innodb_flush_method=O_DIRECT on a SATA SSD with 4096-byte block size. During the test, the call to log_make_checkpoint() in row_import_cleanup() was present. buf_flush_list(): Make static. buf_flush_wait(): Wait for buf_pool.get_oldest_modification() to reach a target, by work done in the buf_flush_page_cleaner. If buf_flush_sync_lsn is going to be set, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_ahead(): If buf_flush_sync_lsn or buf_flush_async_lsn is going to be set and the page cleaner woken up, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_wait_flushed(): Invoke buf_flush_wait(). buf_flush_sync(): Invoke recv_sys.apply() at the start in case crash recovery is active. Invoke buf_flush_wait(). buf_flush_sync_batch(): A lower-level variant of buf_flush_sync() that is only called by recv_sys_t::apply(). buf_flush_sync_for_checkpoint(): Do not trigger log apply or checkpoint during recovery. buf_dblwr_t::create(): Only initiate a buffer pool flush, not a checkpoint. row_import_cleanup(): Do not unnecessarily invoke log_make_checkpoint(). Invoking buf_flush_list_space() before starting to generate redo log for the imported tablespace should suffice. srv_prepare_to_delete_redo_log_file(): Set recv_sys.recovery_on in order to prevent buf_flush_sync_for_checkpoint() from initiating a checkpoint while the log is inaccessible. Remove a wait loop that is already part of buf_flush_sync(). Do not invoke fil_names_clear() if the log is being upgraded, because the FILE_MODIFY record is specific to the latest format. create_log_file(): Clear recv_sys.recovery_on only after calling log_make_checkpoint(), to prevent buf_flush_page_cleaner from invoking a checkpoint. innodb_shutdown(): Simplify the logic in mariadb-backup --prepare. os_aio_wait_until_no_pending_writes(): Update the function comment. Apart from row_quiesce_table_start() during FLUSH TABLES...FOR EXPORT, this is being called by buf_flush_list_space(), which is invoked by ALTER TABLE...IMPORT TABLESPACE as well as some encryption operations.
2022-01-04 06:40:31 +01:00
if (!recv_recovery_is_on() &&
checkpoint_lsn > log_sys.last_checkpoint_lsn + SIZE_OF_FILE_CHECKPOINT)
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
{
mysql_mutex_unlock(&buf_pool.flush_list_mutex);
log_checkpoint_low(checkpoint_lsn, newest_lsn);
mysql_mutex_lock(&buf_pool.flush_list_mutex);
measure= buf_pool.get_oldest_modification(LSN_MAX);
}
else
{
MDEV-23855: Use normal mutex for log_sys.mutex, log_sys.flush_order_mutex With an unreasonably small innodb_log_file_size, the page cleaner thread would frequently acquire log_sys.flush_order_mutex and spend a significant portion of CPU time spinning on that mutex when determining the checkpoint LSN.
2020-10-26 15:43:52 +01:00
mysql_mutex_unlock(&log_sys.mutex);
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
if (!measure)
measure= LSN_MAX;
}
MDEV-23855: Use normal mutex for log_sys.mutex, log_sys.flush_order_mutex With an unreasonably small innodb_log_file_size, the page cleaner thread would frequently acquire log_sys.flush_order_mutex and spend a significant portion of CPU time spinning on that mutex when determining the checkpoint LSN.
2020-10-26 15:43:52 +01:00
mysql_mutex_assert_not_owner(&log_sys.mutex);
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
/* After attempting log checkpoint, check if we have reached our target. */
const lsn_t target= buf_flush_sync_lsn;
if (measure >= target)
buf_flush_sync_lsn= 0;
MDEV-25113: Introduce a page cleaner mode before 'furious flush' MDEV-23855 changed the way how the page cleaner is signaled by user threads. If a threshold is exceeded, a mini-transaction commit would invoke buf_flush_ahead() in order to initiate page flushing before all writers would eventually grind to halt in log_free_check(), waiting for the checkpoint age to reduce. However, buf_flush_ahead() would always initiate 'furious flushing', making the buf_flush_page_cleaner thread write innodb_io_capacity_max pages per batch, and sleeping no time between batches, until the limit LSN is reached. Because this could saturate the I/O subsystem, system throughput could significantly reduce during these 'furious flushing' spikes. With this change, we introduce a gentler version of flush-ahead, which would write innodb_io_capacity_max pages per second until the 'soft limit' is reached. buf_flush_ahead(): Add a parameter to specify whether furious flushing is requested. buf_flush_async_lsn: Similar to buf_flush_sync_lsn, a limit for the less intrusive flushing. buf_flush_page_cleaner(): Keep working until buf_flush_async_lsn has been reached. log_close(): Suppress a warning message in the event that a new log is being created during startup, when old logs did not exist. Return what type of page cleaning will be needed. mtr_t::finish_write(): Also when m_log.is_small(), invoke log_close(). Return what type of page cleaning will be needed. mtr_t::commit(): Invoke buf_flush_ahead() based on the return value of mtr_t::finish_write().
2021-06-23 12:13:16 +02:00
else if (measure >= buf_flush_async_lsn)
buf_flush_async_lsn= 0;
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
MDEV-27416 InnoDB hang in buf_flush_wait_flushed(), on log checkpoint InnoDB could sometimes hang when triggering a log checkpoint. This is due to commit 7b1252c03d7131754d9503560fe507b33ca1f8b4 (MDEV-24278), which introduced an untimed wait to buf_flush_page_cleaner(). The hang was noticed by occasional failures of IMPORT TABLESPACE tests, such as innodb.innodb-wl5522, which would (unnecessarily) invoke log_make_checkpoint() from row_import_cleanup(). The reason of the hang was that buf_flush_page_cleaner() would enter untimed sleep despite buf_flush_sync_lsn being set. The exact failure scenario is unclear, because buf_flush_sync_lsn should actually be protected by buf_pool.flush_list_mutex. We prevent the hang by invoking buf_pool.page_cleaner_set_idle(false) whenever we are setting buf_flush_sync_lsn and signaling buf_pool.do_flush_list. The bulk of these changes was originally developed as a preparation for MDEV-26827, to invoke buf_flush_list() from fewer threads, and tested on 10.6 by Matthias Leich. This fix was tested by running 100 repetitions of 100 concurrent instances of the test innodb.innodb-wl5522 on a RelWithDebInfo build, using ext4fs and innodb_flush_method=O_DIRECT on a SATA SSD with 4096-byte block size. During the test, the call to log_make_checkpoint() in row_import_cleanup() was present. buf_flush_list(): Make static. buf_flush_wait(): Wait for buf_pool.get_oldest_modification() to reach a target, by work done in the buf_flush_page_cleaner. If buf_flush_sync_lsn is going to be set, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_ahead(): If buf_flush_sync_lsn or buf_flush_async_lsn is going to be set and the page cleaner woken up, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_wait_flushed(): Invoke buf_flush_wait(). buf_flush_sync(): Invoke recv_sys.apply() at the start in case crash recovery is active. Invoke buf_flush_wait(). buf_flush_sync_batch(): A lower-level variant of buf_flush_sync() that is only called by recv_sys_t::apply(). buf_flush_sync_for_checkpoint(): Do not trigger log apply or checkpoint during recovery. buf_dblwr_t::create(): Only initiate a buffer pool flush, not a checkpoint. row_import_cleanup(): Do not unnecessarily invoke log_make_checkpoint(). Invoking buf_flush_list_space() before starting to generate redo log for the imported tablespace should suffice. srv_prepare_to_delete_redo_log_file(): Set recv_sys.recovery_on in order to prevent buf_flush_sync_for_checkpoint() from initiating a checkpoint while the log is inaccessible. Remove a wait loop that is already part of buf_flush_sync(). Do not invoke fil_names_clear() if the log is being upgraded, because the FILE_MODIFY record is specific to the latest format. create_log_file(): Clear recv_sys.recovery_on only after calling log_make_checkpoint(), to prevent buf_flush_page_cleaner from invoking a checkpoint. innodb_shutdown(): Simplify the logic in mariadb-backup --prepare. os_aio_wait_until_no_pending_writes(): Update the function comment. Apart from row_quiesce_table_start() during FLUSH TABLES...FOR EXPORT, this is being called by buf_flush_list_space(), which is invoked by ALTER TABLE...IMPORT TABLESPACE as well as some encryption operations.
2022-01-04 06:40:31 +01:00
/* wake up buf_flush_wait() */
MDEV-23399 fixup: Use plain pthread_cond The condition variables that were introduced in commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) are never instrumented with PERFORMANCE_SCHEMA. Let us avoid the storage overhead and dead code.
2021-02-07 11:19:24 +01:00
pthread_cond_broadcast(&buf_pool.done_flush_list);
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
lsn= std::max(lsn, target);
if (measure >= lsn)
return;
}
}
MDEV-25093: Adaptive flushing fails to kick in even if innodb_adaptive_flushing_lwm is hit. (possible regression) adaptive flushing should kick in if a. dirty_pct (dirty pages in buffer pool) > innodb_max_dirty_pages_pct_lwm OR b. innodb_adaptive_flushing_lwm limit is reached (default to 10%) both conditions are mutually exclusive and whichever is first to evaluate true should kick-start the adaptive flushing. After recent changes to simplify the flushing algorithm logic, (b) got ignored that introduced the said regression.
2021-03-29 07:14:02 +02:00
/** Check if the adpative flushing threshold is recommended based on
redo log capacity filled threshold.
@param oldest_lsn buf_pool.get_oldest_modification()
@return true if adaptive flushing is recommended. */
static bool af_needed_for_redo(lsn_t oldest_lsn)
{
lsn_t age= (log_sys.get_lsn() - oldest_lsn);
lsn_t af_lwm= static_cast<lsn_t>(srv_adaptive_flushing_lwm *
MDEV-25113: Introduce a page cleaner mode before 'furious flush' MDEV-23855 changed the way how the page cleaner is signaled by user threads. If a threshold is exceeded, a mini-transaction commit would invoke buf_flush_ahead() in order to initiate page flushing before all writers would eventually grind to halt in log_free_check(), waiting for the checkpoint age to reduce. However, buf_flush_ahead() would always initiate 'furious flushing', making the buf_flush_page_cleaner thread write innodb_io_capacity_max pages per batch, and sleeping no time between batches, until the limit LSN is reached. Because this could saturate the I/O subsystem, system throughput could significantly reduce during these 'furious flushing' spikes. With this change, we introduce a gentler version of flush-ahead, which would write innodb_io_capacity_max pages per second until the 'soft limit' is reached. buf_flush_ahead(): Add a parameter to specify whether furious flushing is requested. buf_flush_async_lsn: Similar to buf_flush_sync_lsn, a limit for the less intrusive flushing. buf_flush_page_cleaner(): Keep working until buf_flush_async_lsn has been reached. log_close(): Suppress a warning message in the event that a new log is being created during startup, when old logs did not exist. Return what type of page cleaning will be needed. mtr_t::finish_write(): Also when m_log.is_small(), invoke log_close(). Return what type of page cleaning will be needed. mtr_t::commit(): Invoke buf_flush_ahead() based on the return value of mtr_t::finish_write().
2021-06-23 12:13:16 +02:00
static_cast<double>(log_sys.log_capacity) / 100);
MDEV-25093: Adaptive flushing fails to kick in even if innodb_adaptive_flushing_lwm is hit. (possible regression) adaptive flushing should kick in if a. dirty_pct (dirty pages in buffer pool) > innodb_max_dirty_pages_pct_lwm OR b. innodb_adaptive_flushing_lwm limit is reached (default to 10%) both conditions are mutually exclusive and whichever is first to evaluate true should kick-start the adaptive flushing. After recent changes to simplify the flushing algorithm logic, (b) got ignored that introduced the said regression.
2021-03-29 07:14:02 +02:00
/* if age > af_lwm adaptive flushing is recommended */
return (age > af_lwm);
}
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
/*********************************************************************//**
Calculates if flushing is required based on redo generation rate.
@return percent of io_capacity to flush to manage redo space */
static
ulint
af_get_pct_for_lsn(
/*===============*/
lsn_t age) /*!< in: current age of LSN. */
{
MDEV-21907: Fix most clang -Wconversion in InnoDB Declare innodb_purge_threads as 4-byte integer (UINT) instead of 4-or-8-byte (ULONG) and adjust the documentation string.
2020-03-10 19:05:17 +01:00
lsn_t af_lwm = static_cast<lsn_t>(
srv_adaptive_flushing_lwm
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
* static_cast<double>(log_sys.log_capacity) / 100);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
if (age < af_lwm) {
/* No adaptive flushing. */
return(0);
}
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
lsn_t lsn_age_factor = (age * 100) / log_sys.max_modified_age_async;
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
ut_ad(srv_max_io_capacity >= srv_io_capacity);
MDEV-21907: Fix most clang -Wconversion in InnoDB Declare innodb_purge_threads as 4-byte integer (UINT) instead of 4-or-8-byte (ULONG) and adjust the documentation string.
2020-03-10 19:05:17 +01:00
return static_cast<ulint>(
(static_cast<double>(srv_max_io_capacity / srv_io_capacity
* lsn_age_factor)
* sqrt(static_cast<double>(lsn_age_factor))
/ 7.5));
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
}
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
/** This function is called approximately once every second by the
page_cleaner thread if innodb_adaptive_flushing=ON.
Based on various factors it decides if there is a need to do flushing.
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
@return number of pages recommended to be flushed
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
@param last_pages_in number of pages flushed in previous batch
@param oldest_lsn buf_pool.get_oldest_modification(0)
MDEV-24369 Page cleaner sleeps despite innodb_max_dirty_pages_pct_lwm being exceeded MDEV-24278 improved the page cleaner so that it will no longer wake up once per second on an idle server. However, with innodb_adaptive_flushing (the default) the function page_cleaner_flush_pages_recommendation() could initially return 0 even if there is work to do. af_get_pct_for_dirty(): Remove. Based on a comment here, it appears that an initial intention of innodb_max_dirty_pages_pct_lwm=0.0 (the default value) was to disable something. That ceased to hold in MDEV-23855: the value is a pure threshold; the page cleaner will not perform any work unless the threshold is exceeded. page_cleaner_flush_pages_recommendation(): Add the parameter dirty_blocks to ensure that buf_pool.flush_list will eventually be emptied.
2020-12-08 13:57:51 +01:00
@param dirty_blocks UT_LIST_GET_LEN(buf_pool.flush_list)
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
@param dirty_pct 100*flush_list.count / (LRU.count + free.count) */
static ulint page_cleaner_flush_pages_recommendation(ulint last_pages_in,
lsn_t oldest_lsn,
MDEV-24369 Page cleaner sleeps despite innodb_max_dirty_pages_pct_lwm being exceeded MDEV-24278 improved the page cleaner so that it will no longer wake up once per second on an idle server. However, with innodb_adaptive_flushing (the default) the function page_cleaner_flush_pages_recommendation() could initially return 0 even if there is work to do. af_get_pct_for_dirty(): Remove. Based on a comment here, it appears that an initial intention of innodb_max_dirty_pages_pct_lwm=0.0 (the default value) was to disable something. That ceased to hold in MDEV-23855: the value is a pure threshold; the page cleaner will not perform any work unless the threshold is exceeded. page_cleaner_flush_pages_recommendation(): Add the parameter dirty_blocks to ensure that buf_pool.flush_list will eventually be emptied.
2020-12-08 13:57:51 +01:00
ulint dirty_blocks,
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
double dirty_pct)
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
{
static lsn_t prev_lsn = 0;
static ulint sum_pages = 0;
static ulint avg_page_rate = 0;
static ulint n_iterations = 0;
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
static time_t prev_time;
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
lsn_t lsn_rate;
ulint n_pages = 0;
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
const lsn_t cur_lsn = log_sys.get_lsn();
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
ut_ad(oldest_lsn <= cur_lsn);
ulint pct_for_lsn = af_get_pct_for_lsn(cur_lsn - oldest_lsn);
time_t curr_time = time(nullptr);
MDEV-24369 Page cleaner sleeps despite innodb_max_dirty_pages_pct_lwm being exceeded MDEV-24278 improved the page cleaner so that it will no longer wake up once per second on an idle server. However, with innodb_adaptive_flushing (the default) the function page_cleaner_flush_pages_recommendation() could initially return 0 even if there is work to do. af_get_pct_for_dirty(): Remove. Based on a comment here, it appears that an initial intention of innodb_max_dirty_pages_pct_lwm=0.0 (the default value) was to disable something. That ceased to hold in MDEV-23855: the value is a pure threshold; the page cleaner will not perform any work unless the threshold is exceeded. page_cleaner_flush_pages_recommendation(): Add the parameter dirty_blocks to ensure that buf_pool.flush_list will eventually be emptied.
2020-12-08 13:57:51 +01:00
const double max_pct = srv_max_buf_pool_modified_pct;
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
if (!prev_lsn || !pct_for_lsn) {
prev_time = curr_time;
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
prev_lsn = cur_lsn;
MDEV-24369 Page cleaner sleeps despite innodb_max_dirty_pages_pct_lwm being exceeded MDEV-24278 improved the page cleaner so that it will no longer wake up once per second on an idle server. However, with innodb_adaptive_flushing (the default) the function page_cleaner_flush_pages_recommendation() could initially return 0 even if there is work to do. af_get_pct_for_dirty(): Remove. Based on a comment here, it appears that an initial intention of innodb_max_dirty_pages_pct_lwm=0.0 (the default value) was to disable something. That ceased to hold in MDEV-23855: the value is a pure threshold; the page cleaner will not perform any work unless the threshold is exceeded. page_cleaner_flush_pages_recommendation(): Add the parameter dirty_blocks to ensure that buf_pool.flush_list will eventually be emptied.
2020-12-08 13:57:51 +01:00
if (max_pct > 0.0) {
dirty_pct /= max_pct;
}
n_pages = ulint(dirty_pct * double(srv_io_capacity));
if (n_pages < dirty_blocks) {
n_pages= std::min<ulint>(srv_io_capacity, dirty_blocks);
}
return n_pages;
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
}
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
sum_pages += last_pages_in;
double time_elapsed = difftime(curr_time, prev_time);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
/* We update our variables every srv_flushing_avg_loops
iterations to smooth out transition in workload. */
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
if (++n_iterations >= srv_flushing_avg_loops
MDEV-21907: Fix most clang -Wconversion in InnoDB Declare innodb_purge_threads as 4-byte integer (UINT) instead of 4-or-8-byte (ULONG) and adjust the documentation string.
2020-03-10 19:05:17 +01:00
|| time_elapsed >= static_cast<double>(srv_flushing_avg_loops)) {
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
if (time_elapsed < 1) {
time_elapsed = 1;
}
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
avg_page_rate = static_cast<ulint>(
((static_cast<double>(sum_pages)
/ time_elapsed)
MDEV-21907: Fix most clang -Wconversion in InnoDB Declare innodb_purge_threads as 4-byte integer (UINT) instead of 4-or-8-byte (ULONG) and adjust the documentation string.
2020-03-10 19:05:17 +01:00
+ static_cast<double>(avg_page_rate)) / 2);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
/* How much LSN we have generated since last call. */
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
lsn_rate = static_cast<lsn_t>(
static_cast<double>(cur_lsn - prev_lsn)
/ time_elapsed);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
lsn_avg_rate = (lsn_avg_rate + lsn_rate) / 2;
Allocate page_cleaner and page_cleaner.slot[] statically
2018-01-24 09:46:04 +01:00
ulint flush_tm = page_cleaner.flush_time;
ulint flush_pass = page_cleaner.flush_pass;
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
Allocate page_cleaner and page_cleaner.slot[] statically
2018-01-24 09:46:04 +01:00
page_cleaner.flush_time = 0;
page_cleaner.flush_pass = 0;
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
if (flush_pass) {
flush_tm /= flush_pass;
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
}
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
MONITOR_SET(MONITOR_FLUSH_ADAPTIVE_AVG_TIME, flush_tm);
MONITOR_SET(MONITOR_FLUSH_ADAPTIVE_AVG_PASS, flush_pass);
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
prev_lsn = cur_lsn;
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
prev_time = curr_time;
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
n_iterations = 0;
sum_pages = 0;
}
MDEV-24537 innodb_max_dirty_pages_pct_lwm=0 lost its special meaning In commit 3a9a3be1c64b14c05648e87ebe0f1dd96457de41 (MDEV-23855) some previous logic was replaced with the condition dirty_pct < srv_max_dirty_pages_pct_lwm, which caused the default value of the parameter innodb_max_dirty_pages_pct_lwm=0 to lose its special meaning: 'refer to innodb_max_dirty_pages_pct instead'. This implicit special meaning was visible in the function af_get_pct_for_dirty(), which was removed in commit f0c295e2de8c074c2ca72e19ff06e1d0e3ee6d2b (MDEV-24369). page_cleaner_flush_pages_recommendation(): Restore the special meaning that was removed in MDEV-24369. buf_flush_page_cleaner(): If srv_max_dirty_pages_pct_lwm==0.0, refer to srv_max_buf_pool_modified_pct. This fixes the observed performance regression due to excessive page flushing. buf_pool_t::page_cleaner_wakeup(): Revise the wakeup condition. innodb_init(): Do initialize srv_max_io_capacity in Mariabackup. It was previously constantly 0, which caused mariadb-backup --prepare to hang in buf_flush_sync(), making no progress.
2021-01-06 12:53:14 +01:00
const ulint pct_for_dirty = srv_max_dirty_pages_pct_lwm == 0
? (dirty_pct >= max_pct ? 100 : 0)
: static_cast<ulint>
MDEV-24369 Page cleaner sleeps despite innodb_max_dirty_pages_pct_lwm being exceeded MDEV-24278 improved the page cleaner so that it will no longer wake up once per second on an idle server. However, with innodb_adaptive_flushing (the default) the function page_cleaner_flush_pages_recommendation() could initially return 0 even if there is work to do. af_get_pct_for_dirty(): Remove. Based on a comment here, it appears that an initial intention of innodb_max_dirty_pages_pct_lwm=0.0 (the default value) was to disable something. That ceased to hold in MDEV-23855: the value is a pure threshold; the page cleaner will not perform any work unless the threshold is exceeded. page_cleaner_flush_pages_recommendation(): Add the parameter dirty_blocks to ensure that buf_pool.flush_list will eventually be emptied.
2020-12-08 13:57:51 +01:00
(max_pct > 0.0 ? dirty_pct / max_pct : dirty_pct);
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
ulint pct_total = std::max(pct_for_dirty, pct_for_lsn);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
/* Estimate pages to be flushed for the lsn progress */
lsn_t target_lsn = oldest_lsn
MDEV-15058: Deprecate and ignore innodb_buffer_pool_instances Our benchmarking efforts indicate that the reasons for splitting the buf_pool in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 have mostly gone away, possibly as a result of mysql/mysql-server@ce6109ebfdedfdf185e391a0c97dc6d33867ed78 or similar work. Only in one write-heavy benchmark where the working set size is ten times the buffer pool size, the buf_pool->mutex would be less contended with 4 buffer pool instances than with 1 instance, in buf_page_io_complete(). That contention could be alleviated further by making more use of std::atomic and by splitting buf_pool_t::mutex further (MDEV-15053). We will deprecate and ignore the following parameters: innodb_buffer_pool_instances innodb_page_cleaners There will be only one buffer pool and one page cleaner task. In a number of INFORMATION_SCHEMA views, columns that indicated the buffer pool instance will be removed: information_schema.innodb_buffer_page.pool_id information_schema.innodb_buffer_page_lru.pool_id information_schema.innodb_buffer_pool_stats.pool_id information_schema.innodb_cmpmem.buffer_pool_instance information_schema.innodb_cmpmem_reset.buffer_pool_instance
2020-02-12 13:45:21 +01:00
+ lsn_avg_rate * buf_flush_lsn_scan_factor;
ulint pages_for_lsn = 0;
MDEV-24369 Page cleaner sleeps despite innodb_max_dirty_pages_pct_lwm being exceeded MDEV-24278 improved the page cleaner so that it will no longer wake up once per second on an idle server. However, with innodb_adaptive_flushing (the default) the function page_cleaner_flush_pages_recommendation() could initially return 0 even if there is work to do. af_get_pct_for_dirty(): Remove. Based on a comment here, it appears that an initial intention of innodb_max_dirty_pages_pct_lwm=0.0 (the default value) was to disable something. That ceased to hold in MDEV-23855: the value is a pure threshold; the page cleaner will not perform any work unless the threshold is exceeded. page_cleaner_flush_pages_recommendation(): Add the parameter dirty_blocks to ensure that buf_pool.flush_list will eventually be emptied.
2020-12-08 13:57:51 +01:00
mysql_mutex_lock(&buf_pool.flush_list_mutex);
MDEV-21962 Allocate buf_pool statically Thanks to MDEV-15058, there is only one InnoDB buffer pool. Allocating buf_pool statically removes one level of pointer indirection and makes code more readable, and removes the awkward initialization of some buf_pool members. While doing this, we will also declare some buf_pool_t data members private and replace some functions with member functions. This is mostly affecting buffer pool resizing. This is not aiming to be a complete rewrite of buf_pool_t to a proper class. Most of the buffer pool interface, such as buf_page_get_gen(), will remain in the C programming style for now. buf_pool_t::withdrawing: Replaces buf_pool_withdrawing. buf_pool_t::withdraw_clock_: Replaces buf_withdraw_clock. buf_pool_t::create(): Repalces buf_pool_init(). buf_pool_t::close(): Replaces buf_pool_free(). buf_bool_t::will_be_withdrawn(): Replaces buf_block_will_be_withdrawn(), buf_frame_will_be_withdrawn(). buf_pool_t::clear_hash_index(): Replaces buf_pool_clear_hash_index(). buf_pool_t::get_n_pages(): Replaces buf_pool_get_n_pages(). buf_pool_t::validate(): Replaces buf_validate(). buf_pool_t::print(): Replaces buf_print(). buf_pool_t::block_from_ahi(): Replaces buf_block_from_ahi(). buf_pool_t::is_block_field(): Replaces buf_pointer_is_block_field(). buf_pool_t::is_block_mutex(): Replaces buf_pool_is_block_mutex(). buf_pool_t::is_block_lock(): Replaces buf_pool_is_block_lock(). buf_pool_t::is_obsolete(): Replaces buf_pool_is_obsolete(). buf_pool_t::io_buf: Make default-constructible. buf_pool_t::io_buf::create(): Delayed 'constructor' buf_pool_t::io_buf::close(): Early 'destructor' HazardPointer: Make default-constructible. Define all member functions inline, also for derived classes.
2020-03-18 20:48:00 +01:00
for (buf_page_t* b = UT_LIST_GET_LAST(buf_pool.flush_list);
MDEV-15058: Deprecate and ignore innodb_buffer_pool_instances Our benchmarking efforts indicate that the reasons for splitting the buf_pool in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 have mostly gone away, possibly as a result of mysql/mysql-server@ce6109ebfdedfdf185e391a0c97dc6d33867ed78 or similar work. Only in one write-heavy benchmark where the working set size is ten times the buffer pool size, the buf_pool->mutex would be less contended with 4 buffer pool instances than with 1 instance, in buf_page_io_complete(). That contention could be alleviated further by making more use of std::atomic and by splitting buf_pool_t::mutex further (MDEV-15053). We will deprecate and ignore the following parameters: innodb_buffer_pool_instances innodb_page_cleaners There will be only one buffer pool and one page cleaner task. In a number of INFORMATION_SCHEMA views, columns that indicated the buffer pool instance will be removed: information_schema.innodb_buffer_page.pool_id information_schema.innodb_buffer_page_lru.pool_id information_schema.innodb_buffer_pool_stats.pool_id information_schema.innodb_cmpmem.buffer_pool_instance information_schema.innodb_cmpmem_reset.buffer_pool_instance
2020-02-12 13:45:21 +01:00
b != NULL;
b = UT_LIST_GET_PREV(list, b)) {
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
if (b->oldest_modification() > target_lsn) {
MDEV-15058: Deprecate and ignore innodb_buffer_pool_instances Our benchmarking efforts indicate that the reasons for splitting the buf_pool in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 have mostly gone away, possibly as a result of mysql/mysql-server@ce6109ebfdedfdf185e391a0c97dc6d33867ed78 or similar work. Only in one write-heavy benchmark where the working set size is ten times the buffer pool size, the buf_pool->mutex would be less contended with 4 buffer pool instances than with 1 instance, in buf_page_io_complete(). That contention could be alleviated further by making more use of std::atomic and by splitting buf_pool_t::mutex further (MDEV-15053). We will deprecate and ignore the following parameters: innodb_buffer_pool_instances innodb_page_cleaners There will be only one buffer pool and one page cleaner task. In a number of INFORMATION_SCHEMA views, columns that indicated the buffer pool instance will be removed: information_schema.innodb_buffer_page.pool_id information_schema.innodb_buffer_page_lru.pool_id information_schema.innodb_buffer_pool_stats.pool_id information_schema.innodb_cmpmem.buffer_pool_instance information_schema.innodb_cmpmem_reset.buffer_pool_instance
2020-02-12 13:45:21 +01:00
break;
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
}
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
if (++pages_for_lsn >= srv_max_io_capacity) {
break;
}
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
}
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
mysql_mutex_unlock(&buf_pool.flush_list_mutex);
MDEV-15058: Deprecate and ignore innodb_buffer_pool_instances Our benchmarking efforts indicate that the reasons for splitting the buf_pool in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 have mostly gone away, possibly as a result of mysql/mysql-server@ce6109ebfdedfdf185e391a0c97dc6d33867ed78 or similar work. Only in one write-heavy benchmark where the working set size is ten times the buffer pool size, the buf_pool->mutex would be less contended with 4 buffer pool instances than with 1 instance, in buf_page_io_complete(). That contention could be alleviated further by making more use of std::atomic and by splitting buf_pool_t::mutex further (MDEV-15053). We will deprecate and ignore the following parameters: innodb_buffer_pool_instances innodb_page_cleaners There will be only one buffer pool and one page cleaner task. In a number of INFORMATION_SCHEMA views, columns that indicated the buffer pool instance will be removed: information_schema.innodb_buffer_page.pool_id information_schema.innodb_buffer_page_lru.pool_id information_schema.innodb_buffer_pool_stats.pool_id information_schema.innodb_cmpmem.buffer_pool_instance information_schema.innodb_cmpmem_reset.buffer_pool_instance
2020-02-12 13:45:21 +01:00
pages_for_lsn /= buf_flush_lsn_scan_factor;
if (pages_for_lsn < 1) {
pages_for_lsn = 1;
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
}
MDEV-21907: Fix most clang -Wconversion in InnoDB Declare innodb_purge_threads as 4-byte integer (UINT) instead of 4-or-8-byte (ULONG) and adjust the documentation string.
2020-03-10 19:05:17 +01:00
n_pages = (ulint(double(srv_io_capacity) * double(pct_total) / 100.0)
+ avg_page_rate + pages_for_lsn) / 3;
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
if (n_pages > srv_max_io_capacity) {
n_pages = srv_max_io_capacity;
}
MONITOR_SET(MONITOR_FLUSH_N_TO_FLUSH_REQUESTED, n_pages);
MDEV-15058: Deprecate and ignore innodb_buffer_pool_instances Our benchmarking efforts indicate that the reasons for splitting the buf_pool in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 have mostly gone away, possibly as a result of mysql/mysql-server@ce6109ebfdedfdf185e391a0c97dc6d33867ed78 or similar work. Only in one write-heavy benchmark where the working set size is ten times the buffer pool size, the buf_pool->mutex would be less contended with 4 buffer pool instances than with 1 instance, in buf_page_io_complete(). That contention could be alleviated further by making more use of std::atomic and by splitting buf_pool_t::mutex further (MDEV-15053). We will deprecate and ignore the following parameters: innodb_buffer_pool_instances innodb_page_cleaners There will be only one buffer pool and one page cleaner task. In a number of INFORMATION_SCHEMA views, columns that indicated the buffer pool instance will be removed: information_schema.innodb_buffer_page.pool_id information_schema.innodb_buffer_page_lru.pool_id information_schema.innodb_buffer_pool_stats.pool_id information_schema.innodb_cmpmem.buffer_pool_instance information_schema.innodb_cmpmem_reset.buffer_pool_instance
2020-02-12 13:45:21 +01:00
MONITOR_SET(MONITOR_FLUSH_N_TO_FLUSH_BY_AGE, pages_for_lsn);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MONITOR_SET(MONITOR_FLUSH_AVG_PAGE_RATE, avg_page_rate);
MONITOR_SET(MONITOR_FLUSH_LSN_AVG_RATE, lsn_avg_rate);
MONITOR_SET(MONITOR_FLUSH_PCT_FOR_DIRTY, pct_for_dirty);
MONITOR_SET(MONITOR_FLUSH_PCT_FOR_LSN, pct_for_lsn);
return(n_pages);
}
cleanup: os_thread_create -> std::thread
2021-03-17 08:03:06 +01:00
/** page_cleaner thread tasked with flushing dirty pages from the buffer
pools. As of now we'll have only one coordinator. */
static void buf_flush_page_cleaner()
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
{
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
my_thread_init();
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
#ifdef UNIV_PFS_THREAD
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
pfs_register_thread(page_cleaner_thread_key);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
#endif /* UNIV_PFS_THREAD */
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
ut_ad(!srv_read_only_mode);
ut_ad(buf_page_cleaner_is_active);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
ulint last_pages= 0;
timespec abstime;
set_timespec(abstime, 1);
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
mysql_mutex_lock(&buf_pool.flush_list_mutex);
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
lsn_t lsn_limit;
MDEV-24949: Enabling idle flushing (possible regression from MDEV-23855) - Currently page cleaner thread will stop flushing if dirty_pct < innodb_max_dirty_pages_pct_lwm. - If the server is not performing any activity then said resources/time could be used to flush the pending dirty pages and keep buffer pool clean for the next burst of the cycle. This flushing is called idle flushing. - flushing logic underwent a complete revamp in 10.5.7/8 and as part of the revamp idle flushing logic got removed. - New proposed logic of idle flushing is based on updated logic of the page cleaner that will enable idle flushing if - buf page cleaner is idle - there are dirty pages (< innodb_max_dirty_pages_pct_lwm) - server is not performing any activity Logic will kickstart the idle flushing bounded by innodb_io_capacity. (Thanks to Marko Makela for reviewing the patch and idea right from the its inception).
2021-02-23 09:21:30 +01:00
ulint last_activity_count= srv_get_activity_count();
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
for (;;)
{
lsn_limit= buf_flush_sync_lsn;
Merge InnoDB 5.7 from mysql-5.7.14. Contains also: MDEV-10549 mysqld: sql/handler.cc:2692: int handler::ha_index_first(uchar*): Assertion `table_share->tmp_table != NO_TMP_TABLE || m_lock_type != 2' failed. (branch bb-10.2-jan) Unlike MySQL, InnoDB still uses THR_LOCK in MariaDB MDEV-10548 Some of the debug sync waits do not work with InnoDB 5.7 (branch bb-10.2-jan) enable tests that were fixed in MDEV-10549 MDEV-10548 Some of the debug sync waits do not work with InnoDB 5.7 (branch bb-10.2-jan) fix main.innodb_mysql_sync - re-enable online alter for partitioned innodb tables
2016-09-06 08:43:16 +02:00
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
if (UNIV_UNLIKELY(lsn_limit != 0))
{
furious_flush:
MDEV-24109 InnoDB hangs with innodb_flush_sync=OFF MDEV-23855 broke the handling of innodb_flush_sync=OFF. That parameter is supposed to limit the page write rate in case the log capacity is being exceeded and log checkpoints are needed. With this fix, the following should pass: ./mtr --mysqld=--loose-innodb-flush-sync=0 One of our best regression tests for page flushing is encryption.innochecksum. With innodb_page_size=16k and innodb_flush_sync=OFF it would likely hang without this fix. log_sys.last_checkpoint_lsn: Declare as Atomic_relaxed<lsn_t> so that we are allowed to read the value while not holding log_sys.mutex. buf_flush_wait_flushed(): Let the page cleaner perform the flushing also if innodb_flush_sync=OFF. After the page cleaner has completed, perform a checkpoint if it is needed, because buf_flush_sync_for_checkpoint() will not be run if innodb_flush_sync=OFF. buf_flush_ahead(): Simplify the condition. We do not really care whether buf_flush_page_cleaner() is running. buf_flush_page_cleaner(): Evaluate innodb_flush_sync at the low level. If innodb_flush_sync=OFF, rate-limit the batches to innodb_io_capacity_max pages per second. Reviewed by: Vladislav Vaintroub
2020-11-04 15:55:36 +01:00
if (UNIV_LIKELY(srv_flush_sync))
{
buf_flush_sync_for_checkpoint(lsn_limit);
last_pages= 0;
set_timespec(abstime, 1);
continue;
}
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
}
MDEV-24109 InnoDB hangs with innodb_flush_sync=OFF MDEV-23855 broke the handling of innodb_flush_sync=OFF. That parameter is supposed to limit the page write rate in case the log capacity is being exceeded and log checkpoints are needed. With this fix, the following should pass: ./mtr --mysqld=--loose-innodb-flush-sync=0 One of our best regression tests for page flushing is encryption.innochecksum. With innodb_page_size=16k and innodb_flush_sync=OFF it would likely hang without this fix. log_sys.last_checkpoint_lsn: Declare as Atomic_relaxed<lsn_t> so that we are allowed to read the value while not holding log_sys.mutex. buf_flush_wait_flushed(): Let the page cleaner perform the flushing also if innodb_flush_sync=OFF. After the page cleaner has completed, perform a checkpoint if it is needed, because buf_flush_sync_for_checkpoint() will not be run if innodb_flush_sync=OFF. buf_flush_ahead(): Simplify the condition. We do not really care whether buf_flush_page_cleaner() is running. buf_flush_page_cleaner(): Evaluate innodb_flush_sync at the low level. If innodb_flush_sync=OFF, rate-limit the batches to innodb_io_capacity_max pages per second. Reviewed by: Vladislav Vaintroub
2020-11-04 15:55:36 +01:00
else if (srv_shutdown_state > SRV_SHUTDOWN_INITIATED)
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
break;
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-24949: Enabling idle flushing (possible regression from MDEV-23855) - Currently page cleaner thread will stop flushing if dirty_pct < innodb_max_dirty_pages_pct_lwm. - If the server is not performing any activity then said resources/time could be used to flush the pending dirty pages and keep buffer pool clean for the next burst of the cycle. This flushing is called idle flushing. - flushing logic underwent a complete revamp in 10.5.7/8 and as part of the revamp idle flushing logic got removed. - New proposed logic of idle flushing is based on updated logic of the page cleaner that will enable idle flushing if - buf page cleaner is idle - there are dirty pages (< innodb_max_dirty_pages_pct_lwm) - server is not performing any activity Logic will kickstart the idle flushing bounded by innodb_io_capacity. (Thanks to Marko Makela for reviewing the patch and idea right from the its inception).
2021-02-23 09:21:30 +01:00
/* If buf pager cleaner is idle and there is no work
(either dirty pages are all flushed or adaptive flushing
is not enabled) then opt for non-timed wait */
if (buf_pool.page_cleaner_idle() &&
(!UT_LIST_GET_LEN(buf_pool.flush_list) ||
srv_max_dirty_pages_pct_lwm == 0.0))
my_cond_wait(&buf_pool.do_flush_list, &buf_pool.flush_list_mutex.m_mutex);
MDEV-24278 InnoDB page cleaner keeps waking up on idle server The purpose of the InnoDB page cleaner subsystem is to write out modified pages from the buffer pool to data files. When the innodb_max_dirty_pages_pct_lwm is not exceeded or innodb_adaptive_flushing=ON decides not to write out anything, the page cleaner should keep sleeping indefinitely until the state of the system changes: a dirty page is added to the buffer pool such that the page cleaner would no longer be idle. buf_flush_page_cleaner(): Explicitly note when the page cleaner is idle. When that happens, use mysql_cond_wait() instead of mysql_cond_timedwait(). buf_flush_insert_into_flush_list(): Wake up the page cleaner if needed. innodb_max_dirty_pages_pct_update(), innodb_max_dirty_pages_pct_lwm_update(): Wake up the page cleaner just in case. Note: buf_flush_ahead(), buf_flush_wait_flushed() and shutdown are already waking up the page cleaner thread.
2020-11-25 15:09:47 +01:00
else
MDEV-23399 fixup: Use plain pthread_cond The condition variables that were introduced in commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) are never instrumented with PERFORMANCE_SCHEMA. Let us avoid the storage overhead and dead code.
2021-02-07 11:19:24 +01:00
my_cond_timedwait(&buf_pool.do_flush_list,
&buf_pool.flush_list_mutex.m_mutex, &abstime);
MDEV-24278 InnoDB page cleaner keeps waking up on idle server The purpose of the InnoDB page cleaner subsystem is to write out modified pages from the buffer pool to data files. When the innodb_max_dirty_pages_pct_lwm is not exceeded or innodb_adaptive_flushing=ON decides not to write out anything, the page cleaner should keep sleeping indefinitely until the state of the system changes: a dirty page is added to the buffer pool such that the page cleaner would no longer be idle. buf_flush_page_cleaner(): Explicitly note when the page cleaner is idle. When that happens, use mysql_cond_wait() instead of mysql_cond_timedwait(). buf_flush_insert_into_flush_list(): Wake up the page cleaner if needed. innodb_max_dirty_pages_pct_update(), innodb_max_dirty_pages_pct_lwm_update(): Wake up the page cleaner just in case. Note: buf_flush_ahead(), buf_flush_wait_flushed() and shutdown are already waking up the page cleaner thread.
2020-11-25 15:09:47 +01:00
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
set_timespec(abstime, 1);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-25113: Introduce a page cleaner mode before 'furious flush' MDEV-23855 changed the way how the page cleaner is signaled by user threads. If a threshold is exceeded, a mini-transaction commit would invoke buf_flush_ahead() in order to initiate page flushing before all writers would eventually grind to halt in log_free_check(), waiting for the checkpoint age to reduce. However, buf_flush_ahead() would always initiate 'furious flushing', making the buf_flush_page_cleaner thread write innodb_io_capacity_max pages per batch, and sleeping no time between batches, until the limit LSN is reached. Because this could saturate the I/O subsystem, system throughput could significantly reduce during these 'furious flushing' spikes. With this change, we introduce a gentler version of flush-ahead, which would write innodb_io_capacity_max pages per second until the 'soft limit' is reached. buf_flush_ahead(): Add a parameter to specify whether furious flushing is requested. buf_flush_async_lsn: Similar to buf_flush_sync_lsn, a limit for the less intrusive flushing. buf_flush_page_cleaner(): Keep working until buf_flush_async_lsn has been reached. log_close(): Suppress a warning message in the event that a new log is being created during startup, when old logs did not exist. Return what type of page cleaning will be needed. mtr_t::finish_write(): Also when m_log.is_small(), invoke log_close(). Return what type of page cleaning will be needed. mtr_t::commit(): Invoke buf_flush_ahead() based on the return value of mtr_t::finish_write().
2021-06-23 12:13:16 +02:00
lsn_t soft_lsn_limit= buf_flush_async_lsn;
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
lsn_limit= buf_flush_sync_lsn;
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
if (UNIV_UNLIKELY(lsn_limit != 0))
MDEV-24109 InnoDB hangs with innodb_flush_sync=OFF MDEV-23855 broke the handling of innodb_flush_sync=OFF. That parameter is supposed to limit the page write rate in case the log capacity is being exceeded and log checkpoints are needed. With this fix, the following should pass: ./mtr --mysqld=--loose-innodb-flush-sync=0 One of our best regression tests for page flushing is encryption.innochecksum. With innodb_page_size=16k and innodb_flush_sync=OFF it would likely hang without this fix. log_sys.last_checkpoint_lsn: Declare as Atomic_relaxed<lsn_t> so that we are allowed to read the value while not holding log_sys.mutex. buf_flush_wait_flushed(): Let the page cleaner perform the flushing also if innodb_flush_sync=OFF. After the page cleaner has completed, perform a checkpoint if it is needed, because buf_flush_sync_for_checkpoint() will not be run if innodb_flush_sync=OFF. buf_flush_ahead(): Simplify the condition. We do not really care whether buf_flush_page_cleaner() is running. buf_flush_page_cleaner(): Evaluate innodb_flush_sync at the low level. If innodb_flush_sync=OFF, rate-limit the batches to innodb_io_capacity_max pages per second. Reviewed by: Vladislav Vaintroub
2020-11-04 15:55:36 +01:00
{
if (UNIV_LIKELY(srv_flush_sync))
goto furious_flush;
}
else if (srv_shutdown_state > SRV_SHUTDOWN_INITIATED)
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
break;
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
const lsn_t oldest_lsn= buf_pool.get_oldest_modification(0);
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
if (!oldest_lsn)
MDEV-24109 InnoDB hangs with innodb_flush_sync=OFF MDEV-23855 broke the handling of innodb_flush_sync=OFF. That parameter is supposed to limit the page write rate in case the log capacity is being exceeded and log checkpoints are needed. With this fix, the following should pass: ./mtr --mysqld=--loose-innodb-flush-sync=0 One of our best regression tests for page flushing is encryption.innochecksum. With innodb_page_size=16k and innodb_flush_sync=OFF it would likely hang without this fix. log_sys.last_checkpoint_lsn: Declare as Atomic_relaxed<lsn_t> so that we are allowed to read the value while not holding log_sys.mutex. buf_flush_wait_flushed(): Let the page cleaner perform the flushing also if innodb_flush_sync=OFF. After the page cleaner has completed, perform a checkpoint if it is needed, because buf_flush_sync_for_checkpoint() will not be run if innodb_flush_sync=OFF. buf_flush_ahead(): Simplify the condition. We do not really care whether buf_flush_page_cleaner() is running. buf_flush_page_cleaner(): Evaluate innodb_flush_sync at the low level. If innodb_flush_sync=OFF, rate-limit the batches to innodb_io_capacity_max pages per second. Reviewed by: Vladislav Vaintroub
2020-11-04 15:55:36 +01:00
{
if (UNIV_UNLIKELY(lsn_limit != 0))
{
buf_flush_sync_lsn= 0;
MDEV-27416 InnoDB hang in buf_flush_wait_flushed(), on log checkpoint InnoDB could sometimes hang when triggering a log checkpoint. This is due to commit 7b1252c03d7131754d9503560fe507b33ca1f8b4 (MDEV-24278), which introduced an untimed wait to buf_flush_page_cleaner(). The hang was noticed by occasional failures of IMPORT TABLESPACE tests, such as innodb.innodb-wl5522, which would (unnecessarily) invoke log_make_checkpoint() from row_import_cleanup(). The reason of the hang was that buf_flush_page_cleaner() would enter untimed sleep despite buf_flush_sync_lsn being set. The exact failure scenario is unclear, because buf_flush_sync_lsn should actually be protected by buf_pool.flush_list_mutex. We prevent the hang by invoking buf_pool.page_cleaner_set_idle(false) whenever we are setting buf_flush_sync_lsn and signaling buf_pool.do_flush_list. The bulk of these changes was originally developed as a preparation for MDEV-26827, to invoke buf_flush_list() from fewer threads, and tested on 10.6 by Matthias Leich. This fix was tested by running 100 repetitions of 100 concurrent instances of the test innodb.innodb-wl5522 on a RelWithDebInfo build, using ext4fs and innodb_flush_method=O_DIRECT on a SATA SSD with 4096-byte block size. During the test, the call to log_make_checkpoint() in row_import_cleanup() was present. buf_flush_list(): Make static. buf_flush_wait(): Wait for buf_pool.get_oldest_modification() to reach a target, by work done in the buf_flush_page_cleaner. If buf_flush_sync_lsn is going to be set, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_ahead(): If buf_flush_sync_lsn or buf_flush_async_lsn is going to be set and the page cleaner woken up, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_wait_flushed(): Invoke buf_flush_wait(). buf_flush_sync(): Invoke recv_sys.apply() at the start in case crash recovery is active. Invoke buf_flush_wait(). buf_flush_sync_batch(): A lower-level variant of buf_flush_sync() that is only called by recv_sys_t::apply(). buf_flush_sync_for_checkpoint(): Do not trigger log apply or checkpoint during recovery. buf_dblwr_t::create(): Only initiate a buffer pool flush, not a checkpoint. row_import_cleanup(): Do not unnecessarily invoke log_make_checkpoint(). Invoking buf_flush_list_space() before starting to generate redo log for the imported tablespace should suffice. srv_prepare_to_delete_redo_log_file(): Set recv_sys.recovery_on in order to prevent buf_flush_sync_for_checkpoint() from initiating a checkpoint while the log is inaccessible. Remove a wait loop that is already part of buf_flush_sync(). Do not invoke fil_names_clear() if the log is being upgraded, because the FILE_MODIFY record is specific to the latest format. create_log_file(): Clear recv_sys.recovery_on only after calling log_make_checkpoint(), to prevent buf_flush_page_cleaner from invoking a checkpoint. innodb_shutdown(): Simplify the logic in mariadb-backup --prepare. os_aio_wait_until_no_pending_writes(): Update the function comment. Apart from row_quiesce_table_start() during FLUSH TABLES...FOR EXPORT, this is being called by buf_flush_list_space(), which is invoked by ALTER TABLE...IMPORT TABLESPACE as well as some encryption operations.
2022-01-04 06:40:31 +01:00
/* wake up buf_flush_wait() */
MDEV-23399 fixup: Use plain pthread_cond The condition variables that were introduced in commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) are never instrumented with PERFORMANCE_SCHEMA. Let us avoid the storage overhead and dead code.
2021-02-07 11:19:24 +01:00
pthread_cond_broadcast(&buf_pool.done_flush_list);
MDEV-24109 InnoDB hangs with innodb_flush_sync=OFF MDEV-23855 broke the handling of innodb_flush_sync=OFF. That parameter is supposed to limit the page write rate in case the log capacity is being exceeded and log checkpoints are needed. With this fix, the following should pass: ./mtr --mysqld=--loose-innodb-flush-sync=0 One of our best regression tests for page flushing is encryption.innochecksum. With innodb_page_size=16k and innodb_flush_sync=OFF it would likely hang without this fix. log_sys.last_checkpoint_lsn: Declare as Atomic_relaxed<lsn_t> so that we are allowed to read the value while not holding log_sys.mutex. buf_flush_wait_flushed(): Let the page cleaner perform the flushing also if innodb_flush_sync=OFF. After the page cleaner has completed, perform a checkpoint if it is needed, because buf_flush_sync_for_checkpoint() will not be run if innodb_flush_sync=OFF. buf_flush_ahead(): Simplify the condition. We do not really care whether buf_flush_page_cleaner() is running. buf_flush_page_cleaner(): Evaluate innodb_flush_sync at the low level. If innodb_flush_sync=OFF, rate-limit the batches to innodb_io_capacity_max pages per second. Reviewed by: Vladislav Vaintroub
2020-11-04 15:55:36 +01:00
}
MDEV-24278 InnoDB page cleaner keeps waking up on idle server The purpose of the InnoDB page cleaner subsystem is to write out modified pages from the buffer pool to data files. When the innodb_max_dirty_pages_pct_lwm is not exceeded or innodb_adaptive_flushing=ON decides not to write out anything, the page cleaner should keep sleeping indefinitely until the state of the system changes: a dirty page is added to the buffer pool such that the page cleaner would no longer be idle. buf_flush_page_cleaner(): Explicitly note when the page cleaner is idle. When that happens, use mysql_cond_wait() instead of mysql_cond_timedwait(). buf_flush_insert_into_flush_list(): Wake up the page cleaner if needed. innodb_max_dirty_pages_pct_update(), innodb_max_dirty_pages_pct_lwm_update(): Wake up the page cleaner just in case. Note: buf_flush_ahead(), buf_flush_wait_flushed() and shutdown are already waking up the page cleaner thread.
2020-11-25 15:09:47 +01:00
unemployed:
MDEV-25113: Introduce a page cleaner mode before 'furious flush' MDEV-23855 changed the way how the page cleaner is signaled by user threads. If a threshold is exceeded, a mini-transaction commit would invoke buf_flush_ahead() in order to initiate page flushing before all writers would eventually grind to halt in log_free_check(), waiting for the checkpoint age to reduce. However, buf_flush_ahead() would always initiate 'furious flushing', making the buf_flush_page_cleaner thread write innodb_io_capacity_max pages per batch, and sleeping no time between batches, until the limit LSN is reached. Because this could saturate the I/O subsystem, system throughput could significantly reduce during these 'furious flushing' spikes. With this change, we introduce a gentler version of flush-ahead, which would write innodb_io_capacity_max pages per second until the 'soft limit' is reached. buf_flush_ahead(): Add a parameter to specify whether furious flushing is requested. buf_flush_async_lsn: Similar to buf_flush_sync_lsn, a limit for the less intrusive flushing. buf_flush_page_cleaner(): Keep working until buf_flush_async_lsn has been reached. log_close(): Suppress a warning message in the event that a new log is being created during startup, when old logs did not exist. Return what type of page cleaning will be needed. mtr_t::finish_write(): Also when m_log.is_small(), invoke log_close(). Return what type of page cleaning will be needed. mtr_t::commit(): Invoke buf_flush_ahead() based on the return value of mtr_t::finish_write().
2021-06-23 12:13:16 +02:00
buf_flush_async_lsn= 0;
MDEV-24278 InnoDB page cleaner keeps waking up on idle server The purpose of the InnoDB page cleaner subsystem is to write out modified pages from the buffer pool to data files. When the innodb_max_dirty_pages_pct_lwm is not exceeded or innodb_adaptive_flushing=ON decides not to write out anything, the page cleaner should keep sleeping indefinitely until the state of the system changes: a dirty page is added to the buffer pool such that the page cleaner would no longer be idle. buf_flush_page_cleaner(): Explicitly note when the page cleaner is idle. When that happens, use mysql_cond_wait() instead of mysql_cond_timedwait(). buf_flush_insert_into_flush_list(): Wake up the page cleaner if needed. innodb_max_dirty_pages_pct_update(), innodb_max_dirty_pages_pct_lwm_update(): Wake up the page cleaner just in case. Note: buf_flush_ahead(), buf_flush_wait_flushed() and shutdown are already waking up the page cleaner thread.
2020-11-25 15:09:47 +01:00
buf_pool.page_cleaner_set_idle(true);
MDEV-26626 InnoDB fails to advance the log checkpoint buf_flush_page_cleaner(): Always try to advance the log checkpoint, even when no pages were flushed during the latest batch. Maybe, since the previous batch, there was an LRU flush that removed the last dirty pages. Failure to advance the log checkpoint will cause unnecessary work in Mariabackup and on crash recovery.
2021-09-16 19:10:42 +02:00
DBUG_EXECUTE_IF("ib_log_checkpoint_avoid", continue;);
mysql_mutex_unlock(&buf_pool.flush_list_mutex);
MDEV-26626 fixup: Do not advance checkpoint during startup While the redo log is being resized in srv_start(), we must not write checkpoint information to the old log. Thanks to Matthias Leich for noticing this.
2021-09-22 15:40:47 +02:00
if (!recv_recovery_is_on() &&
!srv_startup_is_before_trx_rollback_phase &&
srv_operation == SRV_OPERATION_NORMAL)
MDEV-26626 InnoDB fails to advance the log checkpoint buf_flush_page_cleaner(): Always try to advance the log checkpoint, even when no pages were flushed during the latest batch. Maybe, since the previous batch, there was an LRU flush that removed the last dirty pages. Failure to advance the log checkpoint will cause unnecessary work in Mariabackup and on crash recovery.
2021-09-16 19:10:42 +02:00
log_checkpoint();
mysql_mutex_lock(&buf_pool.flush_list_mutex);
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
continue;
MDEV-24109 InnoDB hangs with innodb_flush_sync=OFF MDEV-23855 broke the handling of innodb_flush_sync=OFF. That parameter is supposed to limit the page write rate in case the log capacity is being exceeded and log checkpoints are needed. With this fix, the following should pass: ./mtr --mysqld=--loose-innodb-flush-sync=0 One of our best regression tests for page flushing is encryption.innochecksum. With innodb_page_size=16k and innodb_flush_sync=OFF it would likely hang without this fix. log_sys.last_checkpoint_lsn: Declare as Atomic_relaxed<lsn_t> so that we are allowed to read the value while not holding log_sys.mutex. buf_flush_wait_flushed(): Let the page cleaner perform the flushing also if innodb_flush_sync=OFF. After the page cleaner has completed, perform a checkpoint if it is needed, because buf_flush_sync_for_checkpoint() will not be run if innodb_flush_sync=OFF. buf_flush_ahead(): Simplify the condition. We do not really care whether buf_flush_page_cleaner() is running. buf_flush_page_cleaner(): Evaluate innodb_flush_sync at the low level. If innodb_flush_sync=OFF, rate-limit the batches to innodb_io_capacity_max pages per second. Reviewed by: Vladislav Vaintroub
2020-11-04 15:55:36 +01:00
}
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
const ulint dirty_blocks= UT_LIST_GET_LEN(buf_pool.flush_list);
ut_ad(dirty_blocks);
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
/* We perform dirty reads of the LRU+free list lengths here.
Division by zero is not possible, because buf_pool.flush_list is
guaranteed to be nonempty, and it is a subset of buf_pool.LRU. */
const double dirty_pct= double(dirty_blocks) * 100.0 /
double(UT_LIST_GET_LEN(buf_pool.LRU) + UT_LIST_GET_LEN(buf_pool.free));
MDEV-24949: Enabling idle flushing (possible regression from MDEV-23855) - Currently page cleaner thread will stop flushing if dirty_pct < innodb_max_dirty_pages_pct_lwm. - If the server is not performing any activity then said resources/time could be used to flush the pending dirty pages and keep buffer pool clean for the next burst of the cycle. This flushing is called idle flushing. - flushing logic underwent a complete revamp in 10.5.7/8 and as part of the revamp idle flushing logic got removed. - New proposed logic of idle flushing is based on updated logic of the page cleaner that will enable idle flushing if - buf page cleaner is idle - there are dirty pages (< innodb_max_dirty_pages_pct_lwm) - server is not performing any activity Logic will kickstart the idle flushing bounded by innodb_io_capacity. (Thanks to Marko Makela for reviewing the patch and idea right from the its inception).
2021-02-23 09:21:30 +01:00
bool idle_flush= false;
MDEV-25113: Introduce a page cleaner mode before 'furious flush' MDEV-23855 changed the way how the page cleaner is signaled by user threads. If a threshold is exceeded, a mini-transaction commit would invoke buf_flush_ahead() in order to initiate page flushing before all writers would eventually grind to halt in log_free_check(), waiting for the checkpoint age to reduce. However, buf_flush_ahead() would always initiate 'furious flushing', making the buf_flush_page_cleaner thread write innodb_io_capacity_max pages per batch, and sleeping no time between batches, until the limit LSN is reached. Because this could saturate the I/O subsystem, system throughput could significantly reduce during these 'furious flushing' spikes. With this change, we introduce a gentler version of flush-ahead, which would write innodb_io_capacity_max pages per second until the 'soft limit' is reached. buf_flush_ahead(): Add a parameter to specify whether furious flushing is requested. buf_flush_async_lsn: Similar to buf_flush_sync_lsn, a limit for the less intrusive flushing. buf_flush_page_cleaner(): Keep working until buf_flush_async_lsn has been reached. log_close(): Suppress a warning message in the event that a new log is being created during startup, when old logs did not exist. Return what type of page cleaning will be needed. mtr_t::finish_write(): Also when m_log.is_small(), invoke log_close(). Return what type of page cleaning will be needed. mtr_t::commit(): Invoke buf_flush_ahead() based on the return value of mtr_t::finish_write().
2021-06-23 12:13:16 +02:00
if (lsn_limit || soft_lsn_limit);
MDEV-25093: Adaptive flushing fails to kick in even if innodb_adaptive_flushing_lwm is hit. (possible regression) adaptive flushing should kick in if a. dirty_pct (dirty pages in buffer pool) > innodb_max_dirty_pages_pct_lwm OR b. innodb_adaptive_flushing_lwm limit is reached (default to 10%) both conditions are mutually exclusive and whichever is first to evaluate true should kick-start the adaptive flushing. After recent changes to simplify the flushing algorithm logic, (b) got ignored that introduced the said regression.
2021-03-29 07:14:02 +02:00
else if (af_needed_for_redo(oldest_lsn));
MDEV-24917 Page cleaner wrongly remains idle commit a9933105938d4d809ea33ac3196e681cd581374f (MDEV-24537) introduced the regression that the page cleaner will keep sleeping even if there is work to do. innodb_max_dirty_pages_pct_update(): Always wake up the page cleaner on any SET GLOBAL innodb_max_dirty_pages_pct= assignment. buf_flush_page_cleaner(): If innodb_max_dirty_pages_pct is nonzero, consult only that parameter when determining whether there is work to do. Else, consult innodb_max_dirty_pages.
2021-02-18 17:20:50 +01:00
else if (srv_max_dirty_pages_pct_lwm != 0.0)
{
MDEV-24949: Enabling idle flushing (possible regression from MDEV-23855) - Currently page cleaner thread will stop flushing if dirty_pct < innodb_max_dirty_pages_pct_lwm. - If the server is not performing any activity then said resources/time could be used to flush the pending dirty pages and keep buffer pool clean for the next burst of the cycle. This flushing is called idle flushing. - flushing logic underwent a complete revamp in 10.5.7/8 and as part of the revamp idle flushing logic got removed. - New proposed logic of idle flushing is based on updated logic of the page cleaner that will enable idle flushing if - buf page cleaner is idle - there are dirty pages (< innodb_max_dirty_pages_pct_lwm) - server is not performing any activity Logic will kickstart the idle flushing bounded by innodb_io_capacity. (Thanks to Marko Makela for reviewing the patch and idea right from the its inception).
2021-02-23 09:21:30 +01:00
const ulint activity_count= srv_get_activity_count();
if (activity_count != last_activity_count)
last_activity_count= activity_count;
else if (buf_pool.page_cleaner_idle() && buf_pool.n_pend_reads == 0)
{
/* reaching here means 3 things:
- last_activity_count == activity_count: suggesting server is idle
(no trx_t::commit activity)
- page cleaner is idle (dirty_pct < srv_max_dirty_pages_pct_lwm)
- there are no pending reads but there are dirty pages to flush */
idle_flush= true;
buf_pool.update_last_activity_count(activity_count);
}
if (!idle_flush && dirty_pct < srv_max_dirty_pages_pct_lwm)
MDEV-24917 Page cleaner wrongly remains idle commit a9933105938d4d809ea33ac3196e681cd581374f (MDEV-24537) introduced the regression that the page cleaner will keep sleeping even if there is work to do. innodb_max_dirty_pages_pct_update(): Always wake up the page cleaner on any SET GLOBAL innodb_max_dirty_pages_pct= assignment. buf_flush_page_cleaner(): If innodb_max_dirty_pages_pct is nonzero, consult only that parameter when determining whether there is work to do. Else, consult innodb_max_dirty_pages.
2021-02-18 17:20:50 +01:00
goto unemployed;
}
MDEV-24537 innodb_max_dirty_pages_pct_lwm=0 lost its special meaning In commit 3a9a3be1c64b14c05648e87ebe0f1dd96457de41 (MDEV-23855) some previous logic was replaced with the condition dirty_pct < srv_max_dirty_pages_pct_lwm, which caused the default value of the parameter innodb_max_dirty_pages_pct_lwm=0 to lose its special meaning: 'refer to innodb_max_dirty_pages_pct instead'. This implicit special meaning was visible in the function af_get_pct_for_dirty(), which was removed in commit f0c295e2de8c074c2ca72e19ff06e1d0e3ee6d2b (MDEV-24369). page_cleaner_flush_pages_recommendation(): Restore the special meaning that was removed in MDEV-24369. buf_flush_page_cleaner(): If srv_max_dirty_pages_pct_lwm==0.0, refer to srv_max_buf_pool_modified_pct. This fixes the observed performance regression due to excessive page flushing. buf_pool_t::page_cleaner_wakeup(): Revise the wakeup condition. innodb_init(): Do initialize srv_max_io_capacity in Mariabackup. It was previously constantly 0, which caused mariadb-backup --prepare to hang in buf_flush_sync(), making no progress.
2021-01-06 12:53:14 +01:00
else if (dirty_pct < srv_max_buf_pool_modified_pct)
goto unemployed;
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
MDEV-24109 InnoDB hangs with innodb_flush_sync=OFF MDEV-23855 broke the handling of innodb_flush_sync=OFF. That parameter is supposed to limit the page write rate in case the log capacity is being exceeded and log checkpoints are needed. With this fix, the following should pass: ./mtr --mysqld=--loose-innodb-flush-sync=0 One of our best regression tests for page flushing is encryption.innochecksum. With innodb_page_size=16k and innodb_flush_sync=OFF it would likely hang without this fix. log_sys.last_checkpoint_lsn: Declare as Atomic_relaxed<lsn_t> so that we are allowed to read the value while not holding log_sys.mutex. buf_flush_wait_flushed(): Let the page cleaner perform the flushing also if innodb_flush_sync=OFF. After the page cleaner has completed, perform a checkpoint if it is needed, because buf_flush_sync_for_checkpoint() will not be run if innodb_flush_sync=OFF. buf_flush_ahead(): Simplify the condition. We do not really care whether buf_flush_page_cleaner() is running. buf_flush_page_cleaner(): Evaluate innodb_flush_sync at the low level. If innodb_flush_sync=OFF, rate-limit the batches to innodb_io_capacity_max pages per second. Reviewed by: Vladislav Vaintroub
2020-11-04 15:55:36 +01:00
if (UNIV_UNLIKELY(lsn_limit != 0) && oldest_lsn >= lsn_limit)
MDEV-25113: Introduce a page cleaner mode before 'furious flush' MDEV-23855 changed the way how the page cleaner is signaled by user threads. If a threshold is exceeded, a mini-transaction commit would invoke buf_flush_ahead() in order to initiate page flushing before all writers would eventually grind to halt in log_free_check(), waiting for the checkpoint age to reduce. However, buf_flush_ahead() would always initiate 'furious flushing', making the buf_flush_page_cleaner thread write innodb_io_capacity_max pages per batch, and sleeping no time between batches, until the limit LSN is reached. Because this could saturate the I/O subsystem, system throughput could significantly reduce during these 'furious flushing' spikes. With this change, we introduce a gentler version of flush-ahead, which would write innodb_io_capacity_max pages per second until the 'soft limit' is reached. buf_flush_ahead(): Add a parameter to specify whether furious flushing is requested. buf_flush_async_lsn: Similar to buf_flush_sync_lsn, a limit for the less intrusive flushing. buf_flush_page_cleaner(): Keep working until buf_flush_async_lsn has been reached. log_close(): Suppress a warning message in the event that a new log is being created during startup, when old logs did not exist. Return what type of page cleaning will be needed. mtr_t::finish_write(): Also when m_log.is_small(), invoke log_close(). Return what type of page cleaning will be needed. mtr_t::commit(): Invoke buf_flush_ahead() based on the return value of mtr_t::finish_write().
2021-06-23 12:13:16 +02:00
lsn_limit= buf_flush_sync_lsn= 0;
if (UNIV_UNLIKELY(soft_lsn_limit != 0) && oldest_lsn >= soft_lsn_limit)
soft_lsn_limit= buf_flush_async_lsn= 0;
MDEV-24109 InnoDB hangs with innodb_flush_sync=OFF MDEV-23855 broke the handling of innodb_flush_sync=OFF. That parameter is supposed to limit the page write rate in case the log capacity is being exceeded and log checkpoints are needed. With this fix, the following should pass: ./mtr --mysqld=--loose-innodb-flush-sync=0 One of our best regression tests for page flushing is encryption.innochecksum. With innodb_page_size=16k and innodb_flush_sync=OFF it would likely hang without this fix. log_sys.last_checkpoint_lsn: Declare as Atomic_relaxed<lsn_t> so that we are allowed to read the value while not holding log_sys.mutex. buf_flush_wait_flushed(): Let the page cleaner perform the flushing also if innodb_flush_sync=OFF. After the page cleaner has completed, perform a checkpoint if it is needed, because buf_flush_sync_for_checkpoint() will not be run if innodb_flush_sync=OFF. buf_flush_ahead(): Simplify the condition. We do not really care whether buf_flush_page_cleaner() is running. buf_flush_page_cleaner(): Evaluate innodb_flush_sync at the low level. If innodb_flush_sync=OFF, rate-limit the batches to innodb_io_capacity_max pages per second. Reviewed by: Vladislav Vaintroub
2020-11-04 15:55:36 +01:00
MDEV-24278 InnoDB page cleaner keeps waking up on idle server The purpose of the InnoDB page cleaner subsystem is to write out modified pages from the buffer pool to data files. When the innodb_max_dirty_pages_pct_lwm is not exceeded or innodb_adaptive_flushing=ON decides not to write out anything, the page cleaner should keep sleeping indefinitely until the state of the system changes: a dirty page is added to the buffer pool such that the page cleaner would no longer be idle. buf_flush_page_cleaner(): Explicitly note when the page cleaner is idle. When that happens, use mysql_cond_wait() instead of mysql_cond_timedwait(). buf_flush_insert_into_flush_list(): Wake up the page cleaner if needed. innodb_max_dirty_pages_pct_update(), innodb_max_dirty_pages_pct_lwm_update(): Wake up the page cleaner just in case. Note: buf_flush_ahead(), buf_flush_wait_flushed() and shutdown are already waking up the page cleaner thread.
2020-11-25 15:09:47 +01:00
buf_pool.page_cleaner_set_idle(false);
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
mysql_mutex_unlock(&buf_pool.flush_list_mutex);
MDEV-25113: Introduce a page cleaner mode before 'furious flush' MDEV-23855 changed the way how the page cleaner is signaled by user threads. If a threshold is exceeded, a mini-transaction commit would invoke buf_flush_ahead() in order to initiate page flushing before all writers would eventually grind to halt in log_free_check(), waiting for the checkpoint age to reduce. However, buf_flush_ahead() would always initiate 'furious flushing', making the buf_flush_page_cleaner thread write innodb_io_capacity_max pages per batch, and sleeping no time between batches, until the limit LSN is reached. Because this could saturate the I/O subsystem, system throughput could significantly reduce during these 'furious flushing' spikes. With this change, we introduce a gentler version of flush-ahead, which would write innodb_io_capacity_max pages per second until the 'soft limit' is reached. buf_flush_ahead(): Add a parameter to specify whether furious flushing is requested. buf_flush_async_lsn: Similar to buf_flush_sync_lsn, a limit for the less intrusive flushing. buf_flush_page_cleaner(): Keep working until buf_flush_async_lsn has been reached. log_close(): Suppress a warning message in the event that a new log is being created during startup, when old logs did not exist. Return what type of page cleaning will be needed. mtr_t::finish_write(): Also when m_log.is_small(), invoke log_close(). Return what type of page cleaning will be needed. mtr_t::commit(): Invoke buf_flush_ahead() based on the return value of mtr_t::finish_write().
2021-06-23 12:13:16 +02:00
if (!lsn_limit)
lsn_limit= soft_lsn_limit;
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
ulint n_flushed;
MDEV-24109 InnoDB hangs with innodb_flush_sync=OFF MDEV-23855 broke the handling of innodb_flush_sync=OFF. That parameter is supposed to limit the page write rate in case the log capacity is being exceeded and log checkpoints are needed. With this fix, the following should pass: ./mtr --mysqld=--loose-innodb-flush-sync=0 One of our best regression tests for page flushing is encryption.innochecksum. With innodb_page_size=16k and innodb_flush_sync=OFF it would likely hang without this fix. log_sys.last_checkpoint_lsn: Declare as Atomic_relaxed<lsn_t> so that we are allowed to read the value while not holding log_sys.mutex. buf_flush_wait_flushed(): Let the page cleaner perform the flushing also if innodb_flush_sync=OFF. After the page cleaner has completed, perform a checkpoint if it is needed, because buf_flush_sync_for_checkpoint() will not be run if innodb_flush_sync=OFF. buf_flush_ahead(): Simplify the condition. We do not really care whether buf_flush_page_cleaner() is running. buf_flush_page_cleaner(): Evaluate innodb_flush_sync at the low level. If innodb_flush_sync=OFF, rate-limit the batches to innodb_io_capacity_max pages per second. Reviewed by: Vladislav Vaintroub
2020-11-04 15:55:36 +01:00
if (UNIV_UNLIKELY(lsn_limit != 0))
{
MDEV-25801: buf_flush_dirty_pages() is very slow In commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) the implementation of buf_flush_dirty_pages() was replaced with a slow one, which would perform excessive scans of the buf_pool.flush_list and make little progress. buf_flush_list(), buf_flush_LRU(): Split from buf_flush_lists(). Vladislav Vaintroub noticed that we will not need to invoke log_flush_task.wait() for the LRU eviction flushing. buf_flush_list_space(): Replaces buf_flush_dirty_pages(). This is like buf_flush_list(), but operating on a single tablespace at a time. Writes at most innodb_io_capacity pages. Returns whether some of the tablespace might remain in the buffer pool.
2021-06-23 11:14:26 +02:00
n_flushed= buf_flush_list(srv_max_io_capacity, lsn_limit);
MDEV-27416 InnoDB hang in buf_flush_wait_flushed(), on log checkpoint InnoDB could sometimes hang when triggering a log checkpoint. This is due to commit 7b1252c03d7131754d9503560fe507b33ca1f8b4 (MDEV-24278), which introduced an untimed wait to buf_flush_page_cleaner(). The hang was noticed by occasional failures of IMPORT TABLESPACE tests, such as innodb.innodb-wl5522, which would (unnecessarily) invoke log_make_checkpoint() from row_import_cleanup(). The reason of the hang was that buf_flush_page_cleaner() would enter untimed sleep despite buf_flush_sync_lsn being set. The exact failure scenario is unclear, because buf_flush_sync_lsn should actually be protected by buf_pool.flush_list_mutex. We prevent the hang by invoking buf_pool.page_cleaner_set_idle(false) whenever we are setting buf_flush_sync_lsn and signaling buf_pool.do_flush_list. The bulk of these changes was originally developed as a preparation for MDEV-26827, to invoke buf_flush_list() from fewer threads, and tested on 10.6 by Matthias Leich. This fix was tested by running 100 repetitions of 100 concurrent instances of the test innodb.innodb-wl5522 on a RelWithDebInfo build, using ext4fs and innodb_flush_method=O_DIRECT on a SATA SSD with 4096-byte block size. During the test, the call to log_make_checkpoint() in row_import_cleanup() was present. buf_flush_list(): Make static. buf_flush_wait(): Wait for buf_pool.get_oldest_modification() to reach a target, by work done in the buf_flush_page_cleaner. If buf_flush_sync_lsn is going to be set, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_ahead(): If buf_flush_sync_lsn or buf_flush_async_lsn is going to be set and the page cleaner woken up, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_wait_flushed(): Invoke buf_flush_wait(). buf_flush_sync(): Invoke recv_sys.apply() at the start in case crash recovery is active. Invoke buf_flush_wait(). buf_flush_sync_batch(): A lower-level variant of buf_flush_sync() that is only called by recv_sys_t::apply(). buf_flush_sync_for_checkpoint(): Do not trigger log apply or checkpoint during recovery. buf_dblwr_t::create(): Only initiate a buffer pool flush, not a checkpoint. row_import_cleanup(): Do not unnecessarily invoke log_make_checkpoint(). Invoking buf_flush_list_space() before starting to generate redo log for the imported tablespace should suffice. srv_prepare_to_delete_redo_log_file(): Set recv_sys.recovery_on in order to prevent buf_flush_sync_for_checkpoint() from initiating a checkpoint while the log is inaccessible. Remove a wait loop that is already part of buf_flush_sync(). Do not invoke fil_names_clear() if the log is being upgraded, because the FILE_MODIFY record is specific to the latest format. create_log_file(): Clear recv_sys.recovery_on only after calling log_make_checkpoint(), to prevent buf_flush_page_cleaner from invoking a checkpoint. innodb_shutdown(): Simplify the logic in mariadb-backup --prepare. os_aio_wait_until_no_pending_writes(): Update the function comment. Apart from row_quiesce_table_start() during FLUSH TABLES...FOR EXPORT, this is being called by buf_flush_list_space(), which is invoked by ALTER TABLE...IMPORT TABLESPACE as well as some encryption operations.
2022-01-04 06:40:31 +01:00
/* wake up buf_flush_wait() */
MDEV-23399 fixup: Use plain pthread_cond The condition variables that were introduced in commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) are never instrumented with PERFORMANCE_SCHEMA. Let us avoid the storage overhead and dead code.
2021-02-07 11:19:24 +01:00
pthread_cond_broadcast(&buf_pool.done_flush_list);
MDEV-24109 InnoDB hangs with innodb_flush_sync=OFF MDEV-23855 broke the handling of innodb_flush_sync=OFF. That parameter is supposed to limit the page write rate in case the log capacity is being exceeded and log checkpoints are needed. With this fix, the following should pass: ./mtr --mysqld=--loose-innodb-flush-sync=0 One of our best regression tests for page flushing is encryption.innochecksum. With innodb_page_size=16k and innodb_flush_sync=OFF it would likely hang without this fix. log_sys.last_checkpoint_lsn: Declare as Atomic_relaxed<lsn_t> so that we are allowed to read the value while not holding log_sys.mutex. buf_flush_wait_flushed(): Let the page cleaner perform the flushing also if innodb_flush_sync=OFF. After the page cleaner has completed, perform a checkpoint if it is needed, because buf_flush_sync_for_checkpoint() will not be run if innodb_flush_sync=OFF. buf_flush_ahead(): Simplify the condition. We do not really care whether buf_flush_page_cleaner() is running. buf_flush_page_cleaner(): Evaluate innodb_flush_sync at the low level. If innodb_flush_sync=OFF, rate-limit the batches to innodb_io_capacity_max pages per second. Reviewed by: Vladislav Vaintroub
2020-11-04 15:55:36 +01:00
goto try_checkpoint;
}
MDEV-24949: Enabling idle flushing (possible regression from MDEV-23855) - Currently page cleaner thread will stop flushing if dirty_pct < innodb_max_dirty_pages_pct_lwm. - If the server is not performing any activity then said resources/time could be used to flush the pending dirty pages and keep buffer pool clean for the next burst of the cycle. This flushing is called idle flushing. - flushing logic underwent a complete revamp in 10.5.7/8 and as part of the revamp idle flushing logic got removed. - New proposed logic of idle flushing is based on updated logic of the page cleaner that will enable idle flushing if - buf page cleaner is idle - there are dirty pages (< innodb_max_dirty_pages_pct_lwm) - server is not performing any activity Logic will kickstart the idle flushing bounded by innodb_io_capacity. (Thanks to Marko Makela for reviewing the patch and idea right from the its inception).
2021-02-23 09:21:30 +01:00
else if (idle_flush || !srv_adaptive_flushing)
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
{
MDEV-25801: buf_flush_dirty_pages() is very slow In commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) the implementation of buf_flush_dirty_pages() was replaced with a slow one, which would perform excessive scans of the buf_pool.flush_list and make little progress. buf_flush_list(), buf_flush_LRU(): Split from buf_flush_lists(). Vladislav Vaintroub noticed that we will not need to invoke log_flush_task.wait() for the LRU eviction flushing. buf_flush_list_space(): Replaces buf_flush_dirty_pages(). This is like buf_flush_list(), but operating on a single tablespace at a time. Writes at most innodb_io_capacity pages. Returns whether some of the tablespace might remain in the buffer pool.
2021-06-23 11:14:26 +02:00
n_flushed= buf_flush_list(srv_io_capacity);
MDEV-24109 InnoDB hangs with innodb_flush_sync=OFF MDEV-23855 broke the handling of innodb_flush_sync=OFF. That parameter is supposed to limit the page write rate in case the log capacity is being exceeded and log checkpoints are needed. With this fix, the following should pass: ./mtr --mysqld=--loose-innodb-flush-sync=0 One of our best regression tests for page flushing is encryption.innochecksum. With innodb_page_size=16k and innodb_flush_sync=OFF it would likely hang without this fix. log_sys.last_checkpoint_lsn: Declare as Atomic_relaxed<lsn_t> so that we are allowed to read the value while not holding log_sys.mutex. buf_flush_wait_flushed(): Let the page cleaner perform the flushing also if innodb_flush_sync=OFF. After the page cleaner has completed, perform a checkpoint if it is needed, because buf_flush_sync_for_checkpoint() will not be run if innodb_flush_sync=OFF. buf_flush_ahead(): Simplify the condition. We do not really care whether buf_flush_page_cleaner() is running. buf_flush_page_cleaner(): Evaluate innodb_flush_sync at the low level. If innodb_flush_sync=OFF, rate-limit the batches to innodb_io_capacity_max pages per second. Reviewed by: Vladislav Vaintroub
2020-11-04 15:55:36 +01:00
try_checkpoint:
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
if (n_flushed)
{
MONITOR_INC_VALUE_CUMULATIVE(MONITOR_FLUSH_BACKGROUND_TOTAL_PAGE,
MONITOR_FLUSH_BACKGROUND_COUNT,
MONITOR_FLUSH_BACKGROUND_PAGES,
n_flushed);
do_checkpoint:
/* The periodic log_checkpoint() call here makes it harder to
reproduce bugs in crash recovery or mariabackup --prepare, or
in code that writes the redo log records. Omitting the call
here should not affect correctness, because log_free_check()
should still be invoking checkpoints when needed. */
DBUG_EXECUTE_IF("ib_log_checkpoint_avoid", goto next;);
if (!recv_recovery_is_on() && srv_operation == SRV_OPERATION_NORMAL)
log_checkpoint();
}
}
else if (ulint n= page_cleaner_flush_pages_recommendation(last_pages,
oldest_lsn,
MDEV-24369 Page cleaner sleeps despite innodb_max_dirty_pages_pct_lwm being exceeded MDEV-24278 improved the page cleaner so that it will no longer wake up once per second on an idle server. However, with innodb_adaptive_flushing (the default) the function page_cleaner_flush_pages_recommendation() could initially return 0 even if there is work to do. af_get_pct_for_dirty(): Remove. Based on a comment here, it appears that an initial intention of innodb_max_dirty_pages_pct_lwm=0.0 (the default value) was to disable something. That ceased to hold in MDEV-23855: the value is a pure threshold; the page cleaner will not perform any work unless the threshold is exceeded. page_cleaner_flush_pages_recommendation(): Add the parameter dirty_blocks to ensure that buf_pool.flush_list will eventually be emptied.
2020-12-08 13:57:51 +01:00
dirty_blocks,
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
dirty_pct))
{
page_cleaner.flush_pass++;
const ulint tm= ut_time_ms();
MDEV-25801: buf_flush_dirty_pages() is very slow In commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) the implementation of buf_flush_dirty_pages() was replaced with a slow one, which would perform excessive scans of the buf_pool.flush_list and make little progress. buf_flush_list(), buf_flush_LRU(): Split from buf_flush_lists(). Vladislav Vaintroub noticed that we will not need to invoke log_flush_task.wait() for the LRU eviction flushing. buf_flush_list_space(): Replaces buf_flush_dirty_pages(). This is like buf_flush_list(), but operating on a single tablespace at a time. Writes at most innodb_io_capacity pages. Returns whether some of the tablespace might remain in the buffer pool.
2021-06-23 11:14:26 +02:00
last_pages= n_flushed= buf_flush_list(n);
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
page_cleaner.flush_time+= ut_time_ms() - tm;
if (n_flushed)
{
MONITOR_INC_VALUE_CUMULATIVE(MONITOR_FLUSH_ADAPTIVE_TOTAL_PAGE,
MONITOR_FLUSH_ADAPTIVE_COUNT,
MONITOR_FLUSH_ADAPTIVE_PAGES,
n_flushed);
goto do_checkpoint;
}
}
MDEV-25113: Introduce a page cleaner mode before 'furious flush' MDEV-23855 changed the way how the page cleaner is signaled by user threads. If a threshold is exceeded, a mini-transaction commit would invoke buf_flush_ahead() in order to initiate page flushing before all writers would eventually grind to halt in log_free_check(), waiting for the checkpoint age to reduce. However, buf_flush_ahead() would always initiate 'furious flushing', making the buf_flush_page_cleaner thread write innodb_io_capacity_max pages per batch, and sleeping no time between batches, until the limit LSN is reached. Because this could saturate the I/O subsystem, system throughput could significantly reduce during these 'furious flushing' spikes. With this change, we introduce a gentler version of flush-ahead, which would write innodb_io_capacity_max pages per second until the 'soft limit' is reached. buf_flush_ahead(): Add a parameter to specify whether furious flushing is requested. buf_flush_async_lsn: Similar to buf_flush_sync_lsn, a limit for the less intrusive flushing. buf_flush_page_cleaner(): Keep working until buf_flush_async_lsn has been reached. log_close(): Suppress a warning message in the event that a new log is being created during startup, when old logs did not exist. Return what type of page cleaning will be needed. mtr_t::finish_write(): Also when m_log.is_small(), invoke log_close(). Return what type of page cleaning will be needed. mtr_t::commit(): Invoke buf_flush_ahead() based on the return value of mtr_t::finish_write().
2021-06-23 12:13:16 +02:00
else if (buf_flush_async_lsn <= oldest_lsn)
MDEV-24278 InnoDB page cleaner keeps waking up on idle server The purpose of the InnoDB page cleaner subsystem is to write out modified pages from the buffer pool to data files. When the innodb_max_dirty_pages_pct_lwm is not exceeded or innodb_adaptive_flushing=ON decides not to write out anything, the page cleaner should keep sleeping indefinitely until the state of the system changes: a dirty page is added to the buffer pool such that the page cleaner would no longer be idle. buf_flush_page_cleaner(): Explicitly note when the page cleaner is idle. When that happens, use mysql_cond_wait() instead of mysql_cond_timedwait(). buf_flush_insert_into_flush_list(): Wake up the page cleaner if needed. innodb_max_dirty_pages_pct_update(), innodb_max_dirty_pages_pct_lwm_update(): Wake up the page cleaner just in case. Note: buf_flush_ahead(), buf_flush_wait_flushed() and shutdown are already waking up the page cleaner thread.
2020-11-25 15:09:47 +01:00
{
mysql_mutex_lock(&buf_pool.flush_list_mutex);
goto unemployed;
}
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
#ifndef DBUG_OFF
next:
#endif /* !DBUG_OFF */
mysql_mutex_lock(&buf_pool.flush_list_mutex);
MDEV-24949: Enabling idle flushing (possible regression from MDEV-23855) - Currently page cleaner thread will stop flushing if dirty_pct < innodb_max_dirty_pages_pct_lwm. - If the server is not performing any activity then said resources/time could be used to flush the pending dirty pages and keep buffer pool clean for the next burst of the cycle. This flushing is called idle flushing. - flushing logic underwent a complete revamp in 10.5.7/8 and as part of the revamp idle flushing logic got removed. - New proposed logic of idle flushing is based on updated logic of the page cleaner that will enable idle flushing if - buf page cleaner is idle - there are dirty pages (< innodb_max_dirty_pages_pct_lwm) - server is not performing any activity Logic will kickstart the idle flushing bounded by innodb_io_capacity. (Thanks to Marko Makela for reviewing the patch and idea right from the its inception).
2021-02-23 09:21:30 +01:00
/* when idle flushing kicks in page_cleaner is marked active.
reset it back to idle since the it was made active as part of
idle flushing stage. */
if (idle_flush)
buf_pool.page_cleaner_set_idle(true);
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
}
mysql_mutex_unlock(&buf_pool.flush_list_mutex);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
if (srv_fast_shutdown != 2)
{
buf_flush_wait_batch_end_acquiring_mutex(true);
buf_flush_wait_batch_end_acquiring_mutex(false);
}
mysql_mutex_lock(&buf_pool.flush_list_mutex);
lsn_limit= buf_flush_sync_lsn;
if (UNIV_UNLIKELY(lsn_limit != 0))
goto furious_flush;
buf_page_cleaner_is_active= false;
MDEV-23399 fixup: Use plain pthread_cond The condition variables that were introduced in commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) are never instrumented with PERFORMANCE_SCHEMA. Let us avoid the storage overhead and dead code.
2021-02-07 11:19:24 +01:00
pthread_cond_broadcast(&buf_pool.done_flush_list);
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
mysql_mutex_unlock(&buf_pool.flush_list_mutex);
my_thread_end();
cleanup: os_thread_create -> std::thread
2021-03-17 08:03:06 +01:00
#ifdef UNIV_PFS_THREAD
pfs_delete_thread();
#endif
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
}
MDEV-16264 Use threadpool for Innodb background work. Almost all threads have gone - the "ticking" threads, that sleep a while then do some work) (srv_monitor_thread, srv_error_monitor_thread, srv_master_thread) were replaced with timers. Some timers are periodic, e.g the "master" timer. - The btr_defragment_thread is also replaced by a timer , which reschedules it self when current defragment "item" needs throttling - the buf_resize_thread and buf_dump_threads are substitutes with tasks Ditto with page cleaner workers. - purge workers threads are not tasks as well, and purge cleaner coordinator is a combination of a task and timer. - All AIO is outsourced to tpool, Innodb just calls thread_pool::submit_io() and provides the callback. - The srv_slot_t was removed, and innodb_debug_sync used in purge is currently not working, and needs reimplementation.
2019-10-29 22:37:12 +01:00
MDEV-15058: Deprecate and ignore innodb_buffer_pool_instances Our benchmarking efforts indicate that the reasons for splitting the buf_pool in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 have mostly gone away, possibly as a result of mysql/mysql-server@ce6109ebfdedfdf185e391a0c97dc6d33867ed78 or similar work. Only in one write-heavy benchmark where the working set size is ten times the buffer pool size, the buf_pool->mutex would be less contended with 4 buffer pool instances than with 1 instance, in buf_page_io_complete(). That contention could be alleviated further by making more use of std::atomic and by splitting buf_pool_t::mutex further (MDEV-15053). We will deprecate and ignore the following parameters: innodb_buffer_pool_instances innodb_page_cleaners There will be only one buffer pool and one page cleaner task. In a number of INFORMATION_SCHEMA views, columns that indicated the buffer pool instance will be removed: information_schema.innodb_buffer_page.pool_id information_schema.innodb_buffer_page_lru.pool_id information_schema.innodb_buffer_pool_stats.pool_id information_schema.innodb_cmpmem.buffer_pool_instance information_schema.innodb_cmpmem_reset.buffer_pool_instance
2020-02-12 13:45:21 +01:00
/** Initialize page_cleaner. */
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
ATTRIBUTE_COLD void buf_flush_page_cleaner_init()
MDEV-11025: Make number of page cleaner threads variable dynamic New test cases innodb-page-cleaners Modified test cases innodb_page_cleaners_basic New function buf_flush_set_page_cleaner_thread_cnt Increase or decrease the amount of page cleaner worker threads. In case of increase this function creates based on current abount and requested amount how many new threads should be created. In case of decrease this function sets up the requested amount of threads and uses is_requested event to signal workers. Then we wait until all new treads are started, old threads that should exit signal is_finished or shutdown has marked that page cleaner should finish. buf_flush_page_cleaner_worker Store current thread id and thread_no and then signal event is_finished. If number of used page cleaner threads decrease we shut down those threads that have thread_no greater or equal than number of page configured page cleaners - 1 (note that there will be always page cleaner coordinator). Before exiting we signal is_finished. New function innodb_page_cleaners_threads_update Update function for innodb-page-cleaners system variable. innobase_start_or_create_for_mysql If more than one page cleaner threads is configured we use new function buf_flush_set_page_cleaner_thread_cnt to set up the requested threads (-1 coordinator).
2017-09-09 10:34:12 +02:00
{
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
ut_ad(!buf_page_cleaner_is_active);
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
ut_ad(srv_operation == SRV_OPERATION_NORMAL ||
srv_operation == SRV_OPERATION_RESTORE ||
srv_operation == SRV_OPERATION_RESTORE_EXPORT);
MDEV-25113: Introduce a page cleaner mode before 'furious flush' MDEV-23855 changed the way how the page cleaner is signaled by user threads. If a threshold is exceeded, a mini-transaction commit would invoke buf_flush_ahead() in order to initiate page flushing before all writers would eventually grind to halt in log_free_check(), waiting for the checkpoint age to reduce. However, buf_flush_ahead() would always initiate 'furious flushing', making the buf_flush_page_cleaner thread write innodb_io_capacity_max pages per batch, and sleeping no time between batches, until the limit LSN is reached. Because this could saturate the I/O subsystem, system throughput could significantly reduce during these 'furious flushing' spikes. With this change, we introduce a gentler version of flush-ahead, which would write innodb_io_capacity_max pages per second until the 'soft limit' is reached. buf_flush_ahead(): Add a parameter to specify whether furious flushing is requested. buf_flush_async_lsn: Similar to buf_flush_sync_lsn, a limit for the less intrusive flushing. buf_flush_page_cleaner(): Keep working until buf_flush_async_lsn has been reached. log_close(): Suppress a warning message in the event that a new log is being created during startup, when old logs did not exist. Return what type of page cleaning will be needed. mtr_t::finish_write(): Also when m_log.is_small(), invoke log_close(). Return what type of page cleaning will be needed. mtr_t::commit(): Invoke buf_flush_ahead() based on the return value of mtr_t::finish_write().
2021-06-23 12:13:16 +02:00
buf_flush_async_lsn= 0;
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
buf_flush_sync_lsn= 0;
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
buf_page_cleaner_is_active= true;
cleanup: os_thread_create -> std::thread
2021-03-17 08:03:06 +01:00
std::thread(buf_flush_page_cleaner).detach();
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
}
MDEV-27416 InnoDB hang in buf_flush_wait_flushed(), on log checkpoint InnoDB could sometimes hang when triggering a log checkpoint. This is due to commit 7b1252c03d7131754d9503560fe507b33ca1f8b4 (MDEV-24278), which introduced an untimed wait to buf_flush_page_cleaner(). The hang was noticed by occasional failures of IMPORT TABLESPACE tests, such as innodb.innodb-wl5522, which would (unnecessarily) invoke log_make_checkpoint() from row_import_cleanup(). The reason of the hang was that buf_flush_page_cleaner() would enter untimed sleep despite buf_flush_sync_lsn being set. The exact failure scenario is unclear, because buf_flush_sync_lsn should actually be protected by buf_pool.flush_list_mutex. We prevent the hang by invoking buf_pool.page_cleaner_set_idle(false) whenever we are setting buf_flush_sync_lsn and signaling buf_pool.do_flush_list. The bulk of these changes was originally developed as a preparation for MDEV-26827, to invoke buf_flush_list() from fewer threads, and tested on 10.6 by Matthias Leich. This fix was tested by running 100 repetitions of 100 concurrent instances of the test innodb.innodb-wl5522 on a RelWithDebInfo build, using ext4fs and innodb_flush_method=O_DIRECT on a SATA SSD with 4096-byte block size. During the test, the call to log_make_checkpoint() in row_import_cleanup() was present. buf_flush_list(): Make static. buf_flush_wait(): Wait for buf_pool.get_oldest_modification() to reach a target, by work done in the buf_flush_page_cleaner. If buf_flush_sync_lsn is going to be set, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_ahead(): If buf_flush_sync_lsn or buf_flush_async_lsn is going to be set and the page cleaner woken up, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_wait_flushed(): Invoke buf_flush_wait(). buf_flush_sync(): Invoke recv_sys.apply() at the start in case crash recovery is active. Invoke buf_flush_wait(). buf_flush_sync_batch(): A lower-level variant of buf_flush_sync() that is only called by recv_sys_t::apply(). buf_flush_sync_for_checkpoint(): Do not trigger log apply or checkpoint during recovery. buf_dblwr_t::create(): Only initiate a buffer pool flush, not a checkpoint. row_import_cleanup(): Do not unnecessarily invoke log_make_checkpoint(). Invoking buf_flush_list_space() before starting to generate redo log for the imported tablespace should suffice. srv_prepare_to_delete_redo_log_file(): Set recv_sys.recovery_on in order to prevent buf_flush_sync_for_checkpoint() from initiating a checkpoint while the log is inaccessible. Remove a wait loop that is already part of buf_flush_sync(). Do not invoke fil_names_clear() if the log is being upgraded, because the FILE_MODIFY record is specific to the latest format. create_log_file(): Clear recv_sys.recovery_on only after calling log_make_checkpoint(), to prevent buf_flush_page_cleaner from invoking a checkpoint. innodb_shutdown(): Simplify the logic in mariadb-backup --prepare. os_aio_wait_until_no_pending_writes(): Update the function comment. Apart from row_quiesce_table_start() during FLUSH TABLES...FOR EXPORT, this is being called by buf_flush_list_space(), which is invoked by ALTER TABLE...IMPORT TABLESPACE as well as some encryption operations.
2022-01-04 06:40:31 +01:00
#if defined(HAVE_SYSTEMD) && !defined(EMBEDDED_LIBRARY)
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
/** @return the number of dirty pages in the buffer pool */
static ulint buf_flush_list_length()
{
mysql_mutex_lock(&buf_pool.flush_list_mutex);
const ulint len= UT_LIST_GET_LEN(buf_pool.flush_list);
mysql_mutex_unlock(&buf_pool.flush_list_mutex);
return len;
}
MDEV-27416 InnoDB hang in buf_flush_wait_flushed(), on log checkpoint InnoDB could sometimes hang when triggering a log checkpoint. This is due to commit 7b1252c03d7131754d9503560fe507b33ca1f8b4 (MDEV-24278), which introduced an untimed wait to buf_flush_page_cleaner(). The hang was noticed by occasional failures of IMPORT TABLESPACE tests, such as innodb.innodb-wl5522, which would (unnecessarily) invoke log_make_checkpoint() from row_import_cleanup(). The reason of the hang was that buf_flush_page_cleaner() would enter untimed sleep despite buf_flush_sync_lsn being set. The exact failure scenario is unclear, because buf_flush_sync_lsn should actually be protected by buf_pool.flush_list_mutex. We prevent the hang by invoking buf_pool.page_cleaner_set_idle(false) whenever we are setting buf_flush_sync_lsn and signaling buf_pool.do_flush_list. The bulk of these changes was originally developed as a preparation for MDEV-26827, to invoke buf_flush_list() from fewer threads, and tested on 10.6 by Matthias Leich. This fix was tested by running 100 repetitions of 100 concurrent instances of the test innodb.innodb-wl5522 on a RelWithDebInfo build, using ext4fs and innodb_flush_method=O_DIRECT on a SATA SSD with 4096-byte block size. During the test, the call to log_make_checkpoint() in row_import_cleanup() was present. buf_flush_list(): Make static. buf_flush_wait(): Wait for buf_pool.get_oldest_modification() to reach a target, by work done in the buf_flush_page_cleaner. If buf_flush_sync_lsn is going to be set, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_ahead(): If buf_flush_sync_lsn or buf_flush_async_lsn is going to be set and the page cleaner woken up, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_wait_flushed(): Invoke buf_flush_wait(). buf_flush_sync(): Invoke recv_sys.apply() at the start in case crash recovery is active. Invoke buf_flush_wait(). buf_flush_sync_batch(): A lower-level variant of buf_flush_sync() that is only called by recv_sys_t::apply(). buf_flush_sync_for_checkpoint(): Do not trigger log apply or checkpoint during recovery. buf_dblwr_t::create(): Only initiate a buffer pool flush, not a checkpoint. row_import_cleanup(): Do not unnecessarily invoke log_make_checkpoint(). Invoking buf_flush_list_space() before starting to generate redo log for the imported tablespace should suffice. srv_prepare_to_delete_redo_log_file(): Set recv_sys.recovery_on in order to prevent buf_flush_sync_for_checkpoint() from initiating a checkpoint while the log is inaccessible. Remove a wait loop that is already part of buf_flush_sync(). Do not invoke fil_names_clear() if the log is being upgraded, because the FILE_MODIFY record is specific to the latest format. create_log_file(): Clear recv_sys.recovery_on only after calling log_make_checkpoint(), to prevent buf_flush_page_cleaner from invoking a checkpoint. innodb_shutdown(): Simplify the logic in mariadb-backup --prepare. os_aio_wait_until_no_pending_writes(): Update the function comment. Apart from row_quiesce_table_start() during FLUSH TABLES...FOR EXPORT, this is being called by buf_flush_list_space(), which is invoked by ALTER TABLE...IMPORT TABLESPACE as well as some encryption operations.
2022-01-04 06:40:31 +01:00
#endif
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
/** Flush the buffer pool on shutdown. */
ATTRIBUTE_COLD void buf_flush_buffer_pool()
{
ut_ad(!buf_page_cleaner_is_active);
ut_ad(!buf_flush_sync_lsn);
service_manager_extend_timeout(INNODB_EXTEND_TIMEOUT_INTERVAL,
"Waiting to flush the buffer pool");
MDEV-27416 InnoDB hang in buf_flush_wait_flushed(), on log checkpoint InnoDB could sometimes hang when triggering a log checkpoint. This is due to commit 7b1252c03d7131754d9503560fe507b33ca1f8b4 (MDEV-24278), which introduced an untimed wait to buf_flush_page_cleaner(). The hang was noticed by occasional failures of IMPORT TABLESPACE tests, such as innodb.innodb-wl5522, which would (unnecessarily) invoke log_make_checkpoint() from row_import_cleanup(). The reason of the hang was that buf_flush_page_cleaner() would enter untimed sleep despite buf_flush_sync_lsn being set. The exact failure scenario is unclear, because buf_flush_sync_lsn should actually be protected by buf_pool.flush_list_mutex. We prevent the hang by invoking buf_pool.page_cleaner_set_idle(false) whenever we are setting buf_flush_sync_lsn and signaling buf_pool.do_flush_list. The bulk of these changes was originally developed as a preparation for MDEV-26827, to invoke buf_flush_list() from fewer threads, and tested on 10.6 by Matthias Leich. This fix was tested by running 100 repetitions of 100 concurrent instances of the test innodb.innodb-wl5522 on a RelWithDebInfo build, using ext4fs and innodb_flush_method=O_DIRECT on a SATA SSD with 4096-byte block size. During the test, the call to log_make_checkpoint() in row_import_cleanup() was present. buf_flush_list(): Make static. buf_flush_wait(): Wait for buf_pool.get_oldest_modification() to reach a target, by work done in the buf_flush_page_cleaner. If buf_flush_sync_lsn is going to be set, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_ahead(): If buf_flush_sync_lsn or buf_flush_async_lsn is going to be set and the page cleaner woken up, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_wait_flushed(): Invoke buf_flush_wait(). buf_flush_sync(): Invoke recv_sys.apply() at the start in case crash recovery is active. Invoke buf_flush_wait(). buf_flush_sync_batch(): A lower-level variant of buf_flush_sync() that is only called by recv_sys_t::apply(). buf_flush_sync_for_checkpoint(): Do not trigger log apply or checkpoint during recovery. buf_dblwr_t::create(): Only initiate a buffer pool flush, not a checkpoint. row_import_cleanup(): Do not unnecessarily invoke log_make_checkpoint(). Invoking buf_flush_list_space() before starting to generate redo log for the imported tablespace should suffice. srv_prepare_to_delete_redo_log_file(): Set recv_sys.recovery_on in order to prevent buf_flush_sync_for_checkpoint() from initiating a checkpoint while the log is inaccessible. Remove a wait loop that is already part of buf_flush_sync(). Do not invoke fil_names_clear() if the log is being upgraded, because the FILE_MODIFY record is specific to the latest format. create_log_file(): Clear recv_sys.recovery_on only after calling log_make_checkpoint(), to prevent buf_flush_page_cleaner from invoking a checkpoint. innodb_shutdown(): Simplify the logic in mariadb-backup --prepare. os_aio_wait_until_no_pending_writes(): Update the function comment. Apart from row_quiesce_table_start() during FLUSH TABLES...FOR EXPORT, this is being called by buf_flush_list_space(), which is invoked by ALTER TABLE...IMPORT TABLESPACE as well as some encryption operations.
2022-01-04 06:40:31 +01:00
mysql_mutex_lock(&buf_pool.flush_list_mutex);
while (buf_pool.get_oldest_modification(0))
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
{
MDEV-27416 InnoDB hang in buf_flush_wait_flushed(), on log checkpoint InnoDB could sometimes hang when triggering a log checkpoint. This is due to commit 7b1252c03d7131754d9503560fe507b33ca1f8b4 (MDEV-24278), which introduced an untimed wait to buf_flush_page_cleaner(). The hang was noticed by occasional failures of IMPORT TABLESPACE tests, such as innodb.innodb-wl5522, which would (unnecessarily) invoke log_make_checkpoint() from row_import_cleanup(). The reason of the hang was that buf_flush_page_cleaner() would enter untimed sleep despite buf_flush_sync_lsn being set. The exact failure scenario is unclear, because buf_flush_sync_lsn should actually be protected by buf_pool.flush_list_mutex. We prevent the hang by invoking buf_pool.page_cleaner_set_idle(false) whenever we are setting buf_flush_sync_lsn and signaling buf_pool.do_flush_list. The bulk of these changes was originally developed as a preparation for MDEV-26827, to invoke buf_flush_list() from fewer threads, and tested on 10.6 by Matthias Leich. This fix was tested by running 100 repetitions of 100 concurrent instances of the test innodb.innodb-wl5522 on a RelWithDebInfo build, using ext4fs and innodb_flush_method=O_DIRECT on a SATA SSD with 4096-byte block size. During the test, the call to log_make_checkpoint() in row_import_cleanup() was present. buf_flush_list(): Make static. buf_flush_wait(): Wait for buf_pool.get_oldest_modification() to reach a target, by work done in the buf_flush_page_cleaner. If buf_flush_sync_lsn is going to be set, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_ahead(): If buf_flush_sync_lsn or buf_flush_async_lsn is going to be set and the page cleaner woken up, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_wait_flushed(): Invoke buf_flush_wait(). buf_flush_sync(): Invoke recv_sys.apply() at the start in case crash recovery is active. Invoke buf_flush_wait(). buf_flush_sync_batch(): A lower-level variant of buf_flush_sync() that is only called by recv_sys_t::apply(). buf_flush_sync_for_checkpoint(): Do not trigger log apply or checkpoint during recovery. buf_dblwr_t::create(): Only initiate a buffer pool flush, not a checkpoint. row_import_cleanup(): Do not unnecessarily invoke log_make_checkpoint(). Invoking buf_flush_list_space() before starting to generate redo log for the imported tablespace should suffice. srv_prepare_to_delete_redo_log_file(): Set recv_sys.recovery_on in order to prevent buf_flush_sync_for_checkpoint() from initiating a checkpoint while the log is inaccessible. Remove a wait loop that is already part of buf_flush_sync(). Do not invoke fil_names_clear() if the log is being upgraded, because the FILE_MODIFY record is specific to the latest format. create_log_file(): Clear recv_sys.recovery_on only after calling log_make_checkpoint(), to prevent buf_flush_page_cleaner from invoking a checkpoint. innodb_shutdown(): Simplify the logic in mariadb-backup --prepare. os_aio_wait_until_no_pending_writes(): Update the function comment. Apart from row_quiesce_table_start() during FLUSH TABLES...FOR EXPORT, this is being called by buf_flush_list_space(), which is invoked by ALTER TABLE...IMPORT TABLESPACE as well as some encryption operations.
2022-01-04 06:40:31 +01:00
mysql_mutex_unlock(&buf_pool.flush_list_mutex);
MDEV-25801: buf_flush_dirty_pages() is very slow In commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) the implementation of buf_flush_dirty_pages() was replaced with a slow one, which would perform excessive scans of the buf_pool.flush_list and make little progress. buf_flush_list(), buf_flush_LRU(): Split from buf_flush_lists(). Vladislav Vaintroub noticed that we will not need to invoke log_flush_task.wait() for the LRU eviction flushing. buf_flush_list_space(): Replaces buf_flush_dirty_pages(). This is like buf_flush_list(), but operating on a single tablespace at a time. Writes at most innodb_io_capacity pages. Returns whether some of the tablespace might remain in the buffer pool.
2021-06-23 11:14:26 +02:00
buf_flush_list(srv_max_io_capacity);
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
if (buf_pool.n_flush_list())
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
{
MDEV-27416 InnoDB hang in buf_flush_wait_flushed(), on log checkpoint InnoDB could sometimes hang when triggering a log checkpoint. This is due to commit 7b1252c03d7131754d9503560fe507b33ca1f8b4 (MDEV-24278), which introduced an untimed wait to buf_flush_page_cleaner(). The hang was noticed by occasional failures of IMPORT TABLESPACE tests, such as innodb.innodb-wl5522, which would (unnecessarily) invoke log_make_checkpoint() from row_import_cleanup(). The reason of the hang was that buf_flush_page_cleaner() would enter untimed sleep despite buf_flush_sync_lsn being set. The exact failure scenario is unclear, because buf_flush_sync_lsn should actually be protected by buf_pool.flush_list_mutex. We prevent the hang by invoking buf_pool.page_cleaner_set_idle(false) whenever we are setting buf_flush_sync_lsn and signaling buf_pool.do_flush_list. The bulk of these changes was originally developed as a preparation for MDEV-26827, to invoke buf_flush_list() from fewer threads, and tested on 10.6 by Matthias Leich. This fix was tested by running 100 repetitions of 100 concurrent instances of the test innodb.innodb-wl5522 on a RelWithDebInfo build, using ext4fs and innodb_flush_method=O_DIRECT on a SATA SSD with 4096-byte block size. During the test, the call to log_make_checkpoint() in row_import_cleanup() was present. buf_flush_list(): Make static. buf_flush_wait(): Wait for buf_pool.get_oldest_modification() to reach a target, by work done in the buf_flush_page_cleaner. If buf_flush_sync_lsn is going to be set, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_ahead(): If buf_flush_sync_lsn or buf_flush_async_lsn is going to be set and the page cleaner woken up, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_wait_flushed(): Invoke buf_flush_wait(). buf_flush_sync(): Invoke recv_sys.apply() at the start in case crash recovery is active. Invoke buf_flush_wait(). buf_flush_sync_batch(): A lower-level variant of buf_flush_sync() that is only called by recv_sys_t::apply(). buf_flush_sync_for_checkpoint(): Do not trigger log apply or checkpoint during recovery. buf_dblwr_t::create(): Only initiate a buffer pool flush, not a checkpoint. row_import_cleanup(): Do not unnecessarily invoke log_make_checkpoint(). Invoking buf_flush_list_space() before starting to generate redo log for the imported tablespace should suffice. srv_prepare_to_delete_redo_log_file(): Set recv_sys.recovery_on in order to prevent buf_flush_sync_for_checkpoint() from initiating a checkpoint while the log is inaccessible. Remove a wait loop that is already part of buf_flush_sync(). Do not invoke fil_names_clear() if the log is being upgraded, because the FILE_MODIFY record is specific to the latest format. create_log_file(): Clear recv_sys.recovery_on only after calling log_make_checkpoint(), to prevent buf_flush_page_cleaner from invoking a checkpoint. innodb_shutdown(): Simplify the logic in mariadb-backup --prepare. os_aio_wait_until_no_pending_writes(): Update the function comment. Apart from row_quiesce_table_start() during FLUSH TABLES...FOR EXPORT, this is being called by buf_flush_list_space(), which is invoked by ALTER TABLE...IMPORT TABLESPACE as well as some encryption operations.
2022-01-04 06:40:31 +01:00
timespec abstime;
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
service_manager_extend_timeout(INNODB_EXTEND_TIMEOUT_INTERVAL,
"Waiting to flush " ULINTPF " pages",
buf_flush_list_length());
set_timespec(abstime, INNODB_EXTEND_TIMEOUT_INTERVAL / 2);
mysql_mutex_lock(&buf_pool.mutex);
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
while (buf_pool.n_flush_list_)
MDEV-23399 fixup: Use plain pthread_cond The condition variables that were introduced in commit 7cffb5f6e8a231a041152447be8980ce35d2c9b8 (MDEV-23399) are never instrumented with PERFORMANCE_SCHEMA. Let us avoid the storage overhead and dead code.
2021-02-07 11:19:24 +01:00
my_cond_timedwait(&buf_pool.done_flush_list, &buf_pool.mutex.m_mutex,
&abstime);
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
mysql_mutex_unlock(&buf_pool.mutex);
}
MDEV-27416 InnoDB hang in buf_flush_wait_flushed(), on log checkpoint InnoDB could sometimes hang when triggering a log checkpoint. This is due to commit 7b1252c03d7131754d9503560fe507b33ca1f8b4 (MDEV-24278), which introduced an untimed wait to buf_flush_page_cleaner(). The hang was noticed by occasional failures of IMPORT TABLESPACE tests, such as innodb.innodb-wl5522, which would (unnecessarily) invoke log_make_checkpoint() from row_import_cleanup(). The reason of the hang was that buf_flush_page_cleaner() would enter untimed sleep despite buf_flush_sync_lsn being set. The exact failure scenario is unclear, because buf_flush_sync_lsn should actually be protected by buf_pool.flush_list_mutex. We prevent the hang by invoking buf_pool.page_cleaner_set_idle(false) whenever we are setting buf_flush_sync_lsn and signaling buf_pool.do_flush_list. The bulk of these changes was originally developed as a preparation for MDEV-26827, to invoke buf_flush_list() from fewer threads, and tested on 10.6 by Matthias Leich. This fix was tested by running 100 repetitions of 100 concurrent instances of the test innodb.innodb-wl5522 on a RelWithDebInfo build, using ext4fs and innodb_flush_method=O_DIRECT on a SATA SSD with 4096-byte block size. During the test, the call to log_make_checkpoint() in row_import_cleanup() was present. buf_flush_list(): Make static. buf_flush_wait(): Wait for buf_pool.get_oldest_modification() to reach a target, by work done in the buf_flush_page_cleaner. If buf_flush_sync_lsn is going to be set, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_ahead(): If buf_flush_sync_lsn or buf_flush_async_lsn is going to be set and the page cleaner woken up, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_wait_flushed(): Invoke buf_flush_wait(). buf_flush_sync(): Invoke recv_sys.apply() at the start in case crash recovery is active. Invoke buf_flush_wait(). buf_flush_sync_batch(): A lower-level variant of buf_flush_sync() that is only called by recv_sys_t::apply(). buf_flush_sync_for_checkpoint(): Do not trigger log apply or checkpoint during recovery. buf_dblwr_t::create(): Only initiate a buffer pool flush, not a checkpoint. row_import_cleanup(): Do not unnecessarily invoke log_make_checkpoint(). Invoking buf_flush_list_space() before starting to generate redo log for the imported tablespace should suffice. srv_prepare_to_delete_redo_log_file(): Set recv_sys.recovery_on in order to prevent buf_flush_sync_for_checkpoint() from initiating a checkpoint while the log is inaccessible. Remove a wait loop that is already part of buf_flush_sync(). Do not invoke fil_names_clear() if the log is being upgraded, because the FILE_MODIFY record is specific to the latest format. create_log_file(): Clear recv_sys.recovery_on only after calling log_make_checkpoint(), to prevent buf_flush_page_cleaner from invoking a checkpoint. innodb_shutdown(): Simplify the logic in mariadb-backup --prepare. os_aio_wait_until_no_pending_writes(): Update the function comment. Apart from row_quiesce_table_start() during FLUSH TABLES...FOR EXPORT, this is being called by buf_flush_list_space(), which is invoked by ALTER TABLE...IMPORT TABLESPACE as well as some encryption operations.
2022-01-04 06:40:31 +01:00
mysql_mutex_lock(&buf_pool.flush_list_mutex);
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
}
MDEV-27416 InnoDB hang in buf_flush_wait_flushed(), on log checkpoint InnoDB could sometimes hang when triggering a log checkpoint. This is due to commit 7b1252c03d7131754d9503560fe507b33ca1f8b4 (MDEV-24278), which introduced an untimed wait to buf_flush_page_cleaner(). The hang was noticed by occasional failures of IMPORT TABLESPACE tests, such as innodb.innodb-wl5522, which would (unnecessarily) invoke log_make_checkpoint() from row_import_cleanup(). The reason of the hang was that buf_flush_page_cleaner() would enter untimed sleep despite buf_flush_sync_lsn being set. The exact failure scenario is unclear, because buf_flush_sync_lsn should actually be protected by buf_pool.flush_list_mutex. We prevent the hang by invoking buf_pool.page_cleaner_set_idle(false) whenever we are setting buf_flush_sync_lsn and signaling buf_pool.do_flush_list. The bulk of these changes was originally developed as a preparation for MDEV-26827, to invoke buf_flush_list() from fewer threads, and tested on 10.6 by Matthias Leich. This fix was tested by running 100 repetitions of 100 concurrent instances of the test innodb.innodb-wl5522 on a RelWithDebInfo build, using ext4fs and innodb_flush_method=O_DIRECT on a SATA SSD with 4096-byte block size. During the test, the call to log_make_checkpoint() in row_import_cleanup() was present. buf_flush_list(): Make static. buf_flush_wait(): Wait for buf_pool.get_oldest_modification() to reach a target, by work done in the buf_flush_page_cleaner. If buf_flush_sync_lsn is going to be set, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_ahead(): If buf_flush_sync_lsn or buf_flush_async_lsn is going to be set and the page cleaner woken up, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_wait_flushed(): Invoke buf_flush_wait(). buf_flush_sync(): Invoke recv_sys.apply() at the start in case crash recovery is active. Invoke buf_flush_wait(). buf_flush_sync_batch(): A lower-level variant of buf_flush_sync() that is only called by recv_sys_t::apply(). buf_flush_sync_for_checkpoint(): Do not trigger log apply or checkpoint during recovery. buf_dblwr_t::create(): Only initiate a buffer pool flush, not a checkpoint. row_import_cleanup(): Do not unnecessarily invoke log_make_checkpoint(). Invoking buf_flush_list_space() before starting to generate redo log for the imported tablespace should suffice. srv_prepare_to_delete_redo_log_file(): Set recv_sys.recovery_on in order to prevent buf_flush_sync_for_checkpoint() from initiating a checkpoint while the log is inaccessible. Remove a wait loop that is already part of buf_flush_sync(). Do not invoke fil_names_clear() if the log is being upgraded, because the FILE_MODIFY record is specific to the latest format. create_log_file(): Clear recv_sys.recovery_on only after calling log_make_checkpoint(), to prevent buf_flush_page_cleaner from invoking a checkpoint. innodb_shutdown(): Simplify the logic in mariadb-backup --prepare. os_aio_wait_until_no_pending_writes(): Update the function comment. Apart from row_quiesce_table_start() during FLUSH TABLES...FOR EXPORT, this is being called by buf_flush_list_space(), which is invoked by ALTER TABLE...IMPORT TABLESPACE as well as some encryption operations.
2022-01-04 06:40:31 +01:00
mysql_mutex_unlock(&buf_pool.flush_list_mutex);
MDEV-23855: Improve InnoDB log checkpoint performance After MDEV-15053, MDEV-22871, MDEV-23399 shifted the scalability bottleneck, log checkpoints became a new bottleneck. If innodb_io_capacity is set low or innodb_max_dirty_pct_lwm is set high and the workload fits in the buffer pool, the page cleaner thread will perform very little flushing. When we reach the capacity of the circular redo log file ib_logfile0 and must initiate a checkpoint, some 'furious flushing' will be necessary. (If innodb_flush_sync=OFF, then flushing would continue at the innodb_io_capacity rate, and writers would be throttled.) We have the best chance of advancing the checkpoint LSN immediately after a page flush batch has been completed. Hence, it is best to perform checkpoints after every batch in the page cleaner thread, attempting to run once per second. By initiating high-priority flushing in the page cleaner as early as possible, we aim to make the throughput more stable. The function buf_flush_wait_flushed() used to sleep for 10ms, hoping that the page cleaner thread would do something during that time. The observed end result was that a large number of threads that call log_free_check() would end up sleeping while nothing useful is happening. We will revise the design so that in the default innodb_flush_sync=ON mode, buf_flush_wait_flushed() will wake up the page cleaner thread to perform the necessary flushing, and it will wait for a signal from the page cleaner thread. If innodb_io_capacity is set to a low value (causing the page cleaner to throttle its work), a write workload would initially perform well, until the capacity of the circular ib_logfile0 is reached and log_free_check() will trigger checkpoints. At that point, the extra waiting in buf_flush_wait_flushed() will start reducing throughput. The page cleaner thread will also initiate log checkpoints after each buf_flush_lists() call, because that is the best point of time for the checkpoint LSN to advance by the maximum amount. Even in 'furious flushing' mode we invoke buf_flush_lists() with innodb_io_capacity_max pages at a time, and at the start of each batch (in the log_flush() callback function that runs in a separate task) we will invoke os_aio_wait_until_no_pending_writes(). This tweak allows the checkpoint to advance in smaller steps and significantly reduces the maximum latency. On an Intel Optane 960 NVMe SSD on Linux, it reduced from 4.6 seconds to 74 milliseconds. On Microsoft Windows with a slower SSD, it reduced from more than 180 seconds to 0.6 seconds. We will make innodb_adaptive_flushing=OFF simply flush innodb_io_capacity per second whenever the dirty proportion of buffer pool pages exceeds innodb_max_dirty_pages_pct_lwm. For innodb_adaptive_flushing=ON we try to make page_cleaner_flush_pages_recommendation() more consistent and predictable: if we are below innodb_adaptive_flushing_lwm, let us flush pages according to the return value of af_get_pct_for_dirty(). innodb_max_dirty_pages_pct_lwm: Revert the change of the default value that was made in MDEV-23399. The value innodb_max_dirty_pages_pct_lwm=0 guarantees that a shutdown of an idle server will be fast. Users might be surprised if normal shutdown suddenly became slower when upgrading within a GA release series. innodb_checkpoint_usec: Remove. The master task will no longer perform periodic log checkpoints. It is the duty of the page cleaner thread. log_sys.max_modified_age: Remove. The current span of the buf_pool.flush_list expressed in LSN only matters for adaptive flushing (outside the 'furious flushing' condition). For the correctness of checkpoints, the only thing that matters is the checkpoint age (log_sys.lsn - log_sys.last_checkpoint_lsn). This run-time constant was also reported as log_max_modified_age_sync. log_sys.max_checkpoint_age_async: Remove. This does not serve any purpose, because the checkpoints will now be triggered by the page cleaner thread. We will retain the log_sys.max_checkpoint_age limit for engaging 'furious flushing'. page_cleaner.slot: Remove. It turns out that page_cleaner_slot.flush_list_time was duplicating page_cleaner.slot.flush_time and page_cleaner.slot.flush_list_pass was duplicating page_cleaner.flush_pass. Likewise, there were some redundant monitor counters, because the page cleaner thread no longer performs any buf_pool.LRU flushing, and because there only is one buf_flush_page_cleaner thread. buf_flush_sync_lsn: Protect writes by buf_pool.flush_list_mutex. buf_pool_t::get_oldest_modification(): Add a parameter to specify the return value when no persistent data pages are dirty. Require the caller to hold buf_pool.flush_list_mutex. log_buf_pool_get_oldest_modification(): Take the fall-back LSN as a parameter. All callers will also invoke log_sys.get_lsn(). log_preflush_pool_modified_pages(): Replaced with buf_flush_wait_flushed(). buf_flush_wait_flushed(): Implement two limits. If not enough buffer pool has been flushed, signal the page cleaner (unless innodb_flush_sync=OFF) and wait for the page cleaner to complete. If the page cleaner thread is not running (which can be the case durign shutdown), initiate the flush and wait for it directly. buf_flush_ahead(): If innodb_flush_sync=ON (the default), submit a new buf_flush_sync_lsn target for the page cleaner but do not wait for the flushing to finish. log_get_capacity(), log_get_max_modified_age_async(): Remove, to make it easier to see that af_get_pct_for_lsn() is not acquiring any mutexes. page_cleaner_flush_pages_recommendation(): Protect all access to buf_pool.flush_list with buf_pool.flush_list_mutex. Previously there were some race conditions in the calculation. buf_flush_sync_for_checkpoint(): New function to process buf_flush_sync_lsn in the page cleaner thread. At the end of each batch, we try to wake up any blocked buf_flush_wait_flushed(). If everything up to buf_flush_sync_lsn has been flushed, we will reset buf_flush_sync_lsn=0. The page cleaner thread will keep 'furious flushing' until the limit is reached. Any threads that are waiting in buf_flush_wait_flushed() will be able to resume as soon as their own limit has been satisfied. buf_flush_page_cleaner: Prioritize buf_flush_sync_lsn and do not sleep as long as it is set. Do not update any page_cleaner statistics for this special mode of operation. In the normal mode (buf_flush_sync_lsn is not set for innodb_flush_sync=ON), try to wake up once per second. No longer check whether srv_inc_activity_count() has been called. After each batch, try to perform a log checkpoint, because the best chances for the checkpoint LSN to advance by the maximum amount are upon completing a flushing batch. log_t: Move buf_free, max_buf_free possibly to the same cache line with log_sys.mutex. log_margin_checkpoint_age(): Simplify the logic, and replace a 0.1-second sleep with a call to buf_flush_wait_flushed() to initiate flushing. Moved to the same compilation unit with the only caller. log_close(): Clean up the calculations. (Should be no functional change.) Return whether flush-ahead is needed. Moved to the same compilation unit with the only caller. mtr_t::finish_write(): Return whether flush-ahead is needed. mtr_t::commit(): Invoke buf_flush_ahead() when needed. Let us avoid external calls in mtr_t::commit() and make the logic easier to follow by having related code in a single compilation unit. Also, we will invoke srv_stats.log_write_requests.inc() only once per mini-transaction commit, while not holding mutexes. log_checkpoint_margin(): Only care about log_sys.max_checkpoint_age. Upon reaching log_sys.max_checkpoint_age where we must wait to prevent the log from getting corrupted, let us wait for at most 1MiB of LSN at a time, before rechecking the condition. This should allow writers to proceed even if the redo log capacity has been reached and 'furious flushing' is in progress. We no longer care about log_sys.max_modified_age_sync or log_sys.max_modified_age_async. The log_sys.max_modified_age_sync could be a relic from the time when there was a srv_master_thread that wrote dirty pages to data files. Also, we no longer have any log_sys.max_checkpoint_age_async limit, because log checkpoints will now be triggered by the page cleaner thread upon completing buf_flush_lists(). log_set_capacity(): Simplify the calculations of the limit (no functional change). log_checkpoint_low(): Split from log_checkpoint(). Moved to the same compilation unit with the caller. log_make_checkpoint(): Only wait for everything to be flushed until the current LSN. create_log_file(): After checkpoint, invoke log_write_up_to() to ensure that the FILE_CHECKPOINT record has been written. This avoids ut_ad(!srv_log_file_created) in create_log_file_rename(). srv_start(): Do not call recv_recovery_from_checkpoint_start() if the log has just been created. Set fil_system.space_id_reuse_warned before dict_boot() has been executed, and clear it after recovery has finished. dict_boot(): Initialize fil_system.max_assigned_id. srv_check_activity(): Remove. The activity count is counting transaction commits and therefore mostly interesting for the purge of history. BtrBulk::insert(): Do not explicitly wake up the page cleaner, but do invoke srv_inc_activity_count(), because that counter is still being used in buf_load_throttle_if_needed() for some heuristics. (It might be cleaner to execute buf_load() in the page cleaner thread!) Reviewed by: Vladislav Vaintroub
2020-10-26 15:35:47 +01:00
ut_ad(!buf_pool.any_io_pending());
}
MDEV-27416 InnoDB hang in buf_flush_wait_flushed(), on log checkpoint InnoDB could sometimes hang when triggering a log checkpoint. This is due to commit 7b1252c03d7131754d9503560fe507b33ca1f8b4 (MDEV-24278), which introduced an untimed wait to buf_flush_page_cleaner(). The hang was noticed by occasional failures of IMPORT TABLESPACE tests, such as innodb.innodb-wl5522, which would (unnecessarily) invoke log_make_checkpoint() from row_import_cleanup(). The reason of the hang was that buf_flush_page_cleaner() would enter untimed sleep despite buf_flush_sync_lsn being set. The exact failure scenario is unclear, because buf_flush_sync_lsn should actually be protected by buf_pool.flush_list_mutex. We prevent the hang by invoking buf_pool.page_cleaner_set_idle(false) whenever we are setting buf_flush_sync_lsn and signaling buf_pool.do_flush_list. The bulk of these changes was originally developed as a preparation for MDEV-26827, to invoke buf_flush_list() from fewer threads, and tested on 10.6 by Matthias Leich. This fix was tested by running 100 repetitions of 100 concurrent instances of the test innodb.innodb-wl5522 on a RelWithDebInfo build, using ext4fs and innodb_flush_method=O_DIRECT on a SATA SSD with 4096-byte block size. During the test, the call to log_make_checkpoint() in row_import_cleanup() was present. buf_flush_list(): Make static. buf_flush_wait(): Wait for buf_pool.get_oldest_modification() to reach a target, by work done in the buf_flush_page_cleaner. If buf_flush_sync_lsn is going to be set, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_ahead(): If buf_flush_sync_lsn or buf_flush_async_lsn is going to be set and the page cleaner woken up, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_wait_flushed(): Invoke buf_flush_wait(). buf_flush_sync(): Invoke recv_sys.apply() at the start in case crash recovery is active. Invoke buf_flush_wait(). buf_flush_sync_batch(): A lower-level variant of buf_flush_sync() that is only called by recv_sys_t::apply(). buf_flush_sync_for_checkpoint(): Do not trigger log apply or checkpoint during recovery. buf_dblwr_t::create(): Only initiate a buffer pool flush, not a checkpoint. row_import_cleanup(): Do not unnecessarily invoke log_make_checkpoint(). Invoking buf_flush_list_space() before starting to generate redo log for the imported tablespace should suffice. srv_prepare_to_delete_redo_log_file(): Set recv_sys.recovery_on in order to prevent buf_flush_sync_for_checkpoint() from initiating a checkpoint while the log is inaccessible. Remove a wait loop that is already part of buf_flush_sync(). Do not invoke fil_names_clear() if the log is being upgraded, because the FILE_MODIFY record is specific to the latest format. create_log_file(): Clear recv_sys.recovery_on only after calling log_make_checkpoint(), to prevent buf_flush_page_cleaner from invoking a checkpoint. innodb_shutdown(): Simplify the logic in mariadb-backup --prepare. os_aio_wait_until_no_pending_writes(): Update the function comment. Apart from row_quiesce_table_start() during FLUSH TABLES...FOR EXPORT, this is being called by buf_flush_list_space(), which is invoked by ALTER TABLE...IMPORT TABLESPACE as well as some encryption operations.
2022-01-04 06:40:31 +01:00
/** Synchronously flush dirty blocks during recv_sys_t::apply().
NOTE: The calling thread is not allowed to hold any buffer page latches! */
void buf_flush_sync_batch(lsn_t lsn)
{
thd_wait_begin(nullptr, THD_WAIT_DISKIO);
tpool::tpool_wait_begin();
mysql_mutex_lock(&buf_pool.flush_list_mutex);
buf_flush_wait(lsn);
mysql_mutex_unlock(&buf_pool.flush_list_mutex);
tpool::tpool_wait_end();
thd_wait_end(nullptr);
}
MDEV-15058: Deprecate and ignore innodb_buffer_pool_instances Our benchmarking efforts indicate that the reasons for splitting the buf_pool in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 have mostly gone away, possibly as a result of mysql/mysql-server@ce6109ebfdedfdf185e391a0c97dc6d33867ed78 or similar work. Only in one write-heavy benchmark where the working set size is ten times the buffer pool size, the buf_pool->mutex would be less contended with 4 buffer pool instances than with 1 instance, in buf_page_io_complete(). That contention could be alleviated further by making more use of std::atomic and by splitting buf_pool_t::mutex further (MDEV-15053). We will deprecate and ignore the following parameters: innodb_buffer_pool_instances innodb_page_cleaners There will be only one buffer pool and one page cleaner task. In a number of INFORMATION_SCHEMA views, columns that indicated the buffer pool instance will be removed: information_schema.innodb_buffer_page.pool_id information_schema.innodb_buffer_page_lru.pool_id information_schema.innodb_buffer_pool_stats.pool_id information_schema.innodb_cmpmem.buffer_pool_instance information_schema.innodb_cmpmem_reset.buffer_pool_instance
2020-02-12 13:45:21 +01:00
/** Synchronously flush dirty blocks.
NOTE: The calling thread is not allowed to hold any buffer page latches! */
void buf_flush_sync()
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
{
MDEV-27416 InnoDB hang in buf_flush_wait_flushed(), on log checkpoint InnoDB could sometimes hang when triggering a log checkpoint. This is due to commit 7b1252c03d7131754d9503560fe507b33ca1f8b4 (MDEV-24278), which introduced an untimed wait to buf_flush_page_cleaner(). The hang was noticed by occasional failures of IMPORT TABLESPACE tests, such as innodb.innodb-wl5522, which would (unnecessarily) invoke log_make_checkpoint() from row_import_cleanup(). The reason of the hang was that buf_flush_page_cleaner() would enter untimed sleep despite buf_flush_sync_lsn being set. The exact failure scenario is unclear, because buf_flush_sync_lsn should actually be protected by buf_pool.flush_list_mutex. We prevent the hang by invoking buf_pool.page_cleaner_set_idle(false) whenever we are setting buf_flush_sync_lsn and signaling buf_pool.do_flush_list. The bulk of these changes was originally developed as a preparation for MDEV-26827, to invoke buf_flush_list() from fewer threads, and tested on 10.6 by Matthias Leich. This fix was tested by running 100 repetitions of 100 concurrent instances of the test innodb.innodb-wl5522 on a RelWithDebInfo build, using ext4fs and innodb_flush_method=O_DIRECT on a SATA SSD with 4096-byte block size. During the test, the call to log_make_checkpoint() in row_import_cleanup() was present. buf_flush_list(): Make static. buf_flush_wait(): Wait for buf_pool.get_oldest_modification() to reach a target, by work done in the buf_flush_page_cleaner. If buf_flush_sync_lsn is going to be set, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_ahead(): If buf_flush_sync_lsn or buf_flush_async_lsn is going to be set and the page cleaner woken up, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_wait_flushed(): Invoke buf_flush_wait(). buf_flush_sync(): Invoke recv_sys.apply() at the start in case crash recovery is active. Invoke buf_flush_wait(). buf_flush_sync_batch(): A lower-level variant of buf_flush_sync() that is only called by recv_sys_t::apply(). buf_flush_sync_for_checkpoint(): Do not trigger log apply or checkpoint during recovery. buf_dblwr_t::create(): Only initiate a buffer pool flush, not a checkpoint. row_import_cleanup(): Do not unnecessarily invoke log_make_checkpoint(). Invoking buf_flush_list_space() before starting to generate redo log for the imported tablespace should suffice. srv_prepare_to_delete_redo_log_file(): Set recv_sys.recovery_on in order to prevent buf_flush_sync_for_checkpoint() from initiating a checkpoint while the log is inaccessible. Remove a wait loop that is already part of buf_flush_sync(). Do not invoke fil_names_clear() if the log is being upgraded, because the FILE_MODIFY record is specific to the latest format. create_log_file(): Clear recv_sys.recovery_on only after calling log_make_checkpoint(), to prevent buf_flush_page_cleaner from invoking a checkpoint. innodb_shutdown(): Simplify the logic in mariadb-backup --prepare. os_aio_wait_until_no_pending_writes(): Update the function comment. Apart from row_quiesce_table_start() during FLUSH TABLES...FOR EXPORT, this is being called by buf_flush_list_space(), which is invoked by ALTER TABLE...IMPORT TABLESPACE as well as some encryption operations.
2022-01-04 06:40:31 +01:00
if (recv_recovery_is_on())
recv_sys.apply(true);
thd_wait_begin(nullptr, THD_WAIT_DISKIO);
tpool::tpool_wait_begin();
mysql_mutex_lock(&buf_pool.flush_list_mutex);
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
for (;;)
{
MDEV-27416 InnoDB hang in buf_flush_wait_flushed(), on log checkpoint InnoDB could sometimes hang when triggering a log checkpoint. This is due to commit 7b1252c03d7131754d9503560fe507b33ca1f8b4 (MDEV-24278), which introduced an untimed wait to buf_flush_page_cleaner(). The hang was noticed by occasional failures of IMPORT TABLESPACE tests, such as innodb.innodb-wl5522, which would (unnecessarily) invoke log_make_checkpoint() from row_import_cleanup(). The reason of the hang was that buf_flush_page_cleaner() would enter untimed sleep despite buf_flush_sync_lsn being set. The exact failure scenario is unclear, because buf_flush_sync_lsn should actually be protected by buf_pool.flush_list_mutex. We prevent the hang by invoking buf_pool.page_cleaner_set_idle(false) whenever we are setting buf_flush_sync_lsn and signaling buf_pool.do_flush_list. The bulk of these changes was originally developed as a preparation for MDEV-26827, to invoke buf_flush_list() from fewer threads, and tested on 10.6 by Matthias Leich. This fix was tested by running 100 repetitions of 100 concurrent instances of the test innodb.innodb-wl5522 on a RelWithDebInfo build, using ext4fs and innodb_flush_method=O_DIRECT on a SATA SSD with 4096-byte block size. During the test, the call to log_make_checkpoint() in row_import_cleanup() was present. buf_flush_list(): Make static. buf_flush_wait(): Wait for buf_pool.get_oldest_modification() to reach a target, by work done in the buf_flush_page_cleaner. If buf_flush_sync_lsn is going to be set, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_ahead(): If buf_flush_sync_lsn or buf_flush_async_lsn is going to be set and the page cleaner woken up, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_wait_flushed(): Invoke buf_flush_wait(). buf_flush_sync(): Invoke recv_sys.apply() at the start in case crash recovery is active. Invoke buf_flush_wait(). buf_flush_sync_batch(): A lower-level variant of buf_flush_sync() that is only called by recv_sys_t::apply(). buf_flush_sync_for_checkpoint(): Do not trigger log apply or checkpoint during recovery. buf_dblwr_t::create(): Only initiate a buffer pool flush, not a checkpoint. row_import_cleanup(): Do not unnecessarily invoke log_make_checkpoint(). Invoking buf_flush_list_space() before starting to generate redo log for the imported tablespace should suffice. srv_prepare_to_delete_redo_log_file(): Set recv_sys.recovery_on in order to prevent buf_flush_sync_for_checkpoint() from initiating a checkpoint while the log is inaccessible. Remove a wait loop that is already part of buf_flush_sync(). Do not invoke fil_names_clear() if the log is being upgraded, because the FILE_MODIFY record is specific to the latest format. create_log_file(): Clear recv_sys.recovery_on only after calling log_make_checkpoint(), to prevent buf_flush_page_cleaner from invoking a checkpoint. innodb_shutdown(): Simplify the logic in mariadb-backup --prepare. os_aio_wait_until_no_pending_writes(): Update the function comment. Apart from row_quiesce_table_start() during FLUSH TABLES...FOR EXPORT, this is being called by buf_flush_list_space(), which is invoked by ALTER TABLE...IMPORT TABLESPACE as well as some encryption operations.
2022-01-04 06:40:31 +01:00
const lsn_t lsn= log_sys.get_lsn();
buf_flush_wait(lsn);
MDEV-27499 fixup: Add a wait to buf_flush_sync() The test innodb.log_file_size would occasionally fail with an assertion failure !buf_pool.any_io_pending(). Let us wait for the page cleaner thread to become idle already in srv_prepare_to_delete_redo_log_file(), like we used to.
2022-01-19 18:12:17 +01:00
/* Wait for the page cleaner to be idle (for log resizing at startup) */
while (buf_flush_sync_lsn)
my_cond_wait(&buf_pool.done_flush_list,
&buf_pool.flush_list_mutex.m_mutex);
MDEV-27416 InnoDB hang in buf_flush_wait_flushed(), on log checkpoint InnoDB could sometimes hang when triggering a log checkpoint. This is due to commit 7b1252c03d7131754d9503560fe507b33ca1f8b4 (MDEV-24278), which introduced an untimed wait to buf_flush_page_cleaner(). The hang was noticed by occasional failures of IMPORT TABLESPACE tests, such as innodb.innodb-wl5522, which would (unnecessarily) invoke log_make_checkpoint() from row_import_cleanup(). The reason of the hang was that buf_flush_page_cleaner() would enter untimed sleep despite buf_flush_sync_lsn being set. The exact failure scenario is unclear, because buf_flush_sync_lsn should actually be protected by buf_pool.flush_list_mutex. We prevent the hang by invoking buf_pool.page_cleaner_set_idle(false) whenever we are setting buf_flush_sync_lsn and signaling buf_pool.do_flush_list. The bulk of these changes was originally developed as a preparation for MDEV-26827, to invoke buf_flush_list() from fewer threads, and tested on 10.6 by Matthias Leich. This fix was tested by running 100 repetitions of 100 concurrent instances of the test innodb.innodb-wl5522 on a RelWithDebInfo build, using ext4fs and innodb_flush_method=O_DIRECT on a SATA SSD with 4096-byte block size. During the test, the call to log_make_checkpoint() in row_import_cleanup() was present. buf_flush_list(): Make static. buf_flush_wait(): Wait for buf_pool.get_oldest_modification() to reach a target, by work done in the buf_flush_page_cleaner. If buf_flush_sync_lsn is going to be set, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_ahead(): If buf_flush_sync_lsn or buf_flush_async_lsn is going to be set and the page cleaner woken up, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_wait_flushed(): Invoke buf_flush_wait(). buf_flush_sync(): Invoke recv_sys.apply() at the start in case crash recovery is active. Invoke buf_flush_wait(). buf_flush_sync_batch(): A lower-level variant of buf_flush_sync() that is only called by recv_sys_t::apply(). buf_flush_sync_for_checkpoint(): Do not trigger log apply or checkpoint during recovery. buf_dblwr_t::create(): Only initiate a buffer pool flush, not a checkpoint. row_import_cleanup(): Do not unnecessarily invoke log_make_checkpoint(). Invoking buf_flush_list_space() before starting to generate redo log for the imported tablespace should suffice. srv_prepare_to_delete_redo_log_file(): Set recv_sys.recovery_on in order to prevent buf_flush_sync_for_checkpoint() from initiating a checkpoint while the log is inaccessible. Remove a wait loop that is already part of buf_flush_sync(). Do not invoke fil_names_clear() if the log is being upgraded, because the FILE_MODIFY record is specific to the latest format. create_log_file(): Clear recv_sys.recovery_on only after calling log_make_checkpoint(), to prevent buf_flush_page_cleaner from invoking a checkpoint. innodb_shutdown(): Simplify the logic in mariadb-backup --prepare. os_aio_wait_until_no_pending_writes(): Update the function comment. Apart from row_quiesce_table_start() during FLUSH TABLES...FOR EXPORT, this is being called by buf_flush_list_space(), which is invoked by ALTER TABLE...IMPORT TABLESPACE as well as some encryption operations.
2022-01-04 06:40:31 +01:00
if (lsn == log_sys.get_lsn())
break;
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
}
MDEV-27416 InnoDB hang in buf_flush_wait_flushed(), on log checkpoint InnoDB could sometimes hang when triggering a log checkpoint. This is due to commit 7b1252c03d7131754d9503560fe507b33ca1f8b4 (MDEV-24278), which introduced an untimed wait to buf_flush_page_cleaner(). The hang was noticed by occasional failures of IMPORT TABLESPACE tests, such as innodb.innodb-wl5522, which would (unnecessarily) invoke log_make_checkpoint() from row_import_cleanup(). The reason of the hang was that buf_flush_page_cleaner() would enter untimed sleep despite buf_flush_sync_lsn being set. The exact failure scenario is unclear, because buf_flush_sync_lsn should actually be protected by buf_pool.flush_list_mutex. We prevent the hang by invoking buf_pool.page_cleaner_set_idle(false) whenever we are setting buf_flush_sync_lsn and signaling buf_pool.do_flush_list. The bulk of these changes was originally developed as a preparation for MDEV-26827, to invoke buf_flush_list() from fewer threads, and tested on 10.6 by Matthias Leich. This fix was tested by running 100 repetitions of 100 concurrent instances of the test innodb.innodb-wl5522 on a RelWithDebInfo build, using ext4fs and innodb_flush_method=O_DIRECT on a SATA SSD with 4096-byte block size. During the test, the call to log_make_checkpoint() in row_import_cleanup() was present. buf_flush_list(): Make static. buf_flush_wait(): Wait for buf_pool.get_oldest_modification() to reach a target, by work done in the buf_flush_page_cleaner. If buf_flush_sync_lsn is going to be set, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_ahead(): If buf_flush_sync_lsn or buf_flush_async_lsn is going to be set and the page cleaner woken up, we will invoke buf_pool.page_cleaner_set_idle(false). buf_flush_wait_flushed(): Invoke buf_flush_wait(). buf_flush_sync(): Invoke recv_sys.apply() at the start in case crash recovery is active. Invoke buf_flush_wait(). buf_flush_sync_batch(): A lower-level variant of buf_flush_sync() that is only called by recv_sys_t::apply(). buf_flush_sync_for_checkpoint(): Do not trigger log apply or checkpoint during recovery. buf_dblwr_t::create(): Only initiate a buffer pool flush, not a checkpoint. row_import_cleanup(): Do not unnecessarily invoke log_make_checkpoint(). Invoking buf_flush_list_space() before starting to generate redo log for the imported tablespace should suffice. srv_prepare_to_delete_redo_log_file(): Set recv_sys.recovery_on in order to prevent buf_flush_sync_for_checkpoint() from initiating a checkpoint while the log is inaccessible. Remove a wait loop that is already part of buf_flush_sync(). Do not invoke fil_names_clear() if the log is being upgraded, because the FILE_MODIFY record is specific to the latest format. create_log_file(): Clear recv_sys.recovery_on only after calling log_make_checkpoint(), to prevent buf_flush_page_cleaner from invoking a checkpoint. innodb_shutdown(): Simplify the logic in mariadb-backup --prepare. os_aio_wait_until_no_pending_writes(): Update the function comment. Apart from row_quiesce_table_start() during FLUSH TABLES...FOR EXPORT, this is being called by buf_flush_list_space(), which is invoked by ALTER TABLE...IMPORT TABLESPACE as well as some encryption operations.
2022-01-04 06:40:31 +01:00
mysql_mutex_unlock(&buf_pool.flush_list_mutex);
tpool::tpool_wait_end();
thd_wait_end(nullptr);
Merge InnoDB 5.7 from mysql-5.7.9. Contains also MDEV-10547: Test multi_update_innodb fails with InnoDB 5.7 The failure happened because 5.7 has changed the signature of the bool handler::primary_key_is_clustered() const virtual function ("const" was added). InnoDB was using the old signature which caused the function not to be used. MDEV-10550: Parallel replication lock waits/deadlock handling does not work with InnoDB 5.7 Fixed mutexing problem on lock_trx_handle_wait. Note that rpl_parallel and rpl_optimistic_parallel tests still fail. MDEV-10156 : Group commit tests fail on 10.2 InnoDB (branch bb-10.2-jan) Reason: incorrect merge MDEV-10550: Parallel replication can't sync with master in InnoDB 5.7 (branch bb-10.2-jan) Reason: incorrect merge
2016-08-12 10:17:45 +02:00
}
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
#ifdef UNIV_DEBUG
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
/** Functor to validate the flush list. */
struct Check {
Clean up ut_list ut_list_validate(), ut_list_map(): Add variants with const Functor& so that these functions can be called with an rvalue. Remove wrapper macros, and add #ifdef UNIV_DEBUG around debug-only code.
2019-04-29 13:33:46 +02:00
void operator()(const buf_page_t* elem) const
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
{
MDEV-12227 Defer writes to the InnoDB temporary tablespace The flushing of the InnoDB temporary tablespace is unnecessarily tied to the write-ahead redo logging and redo log checkpoints, which must be tied to the page writes of persistent tablespaces. Let us simply omit any pages of temporary tables from buf_pool.flush_list. In this way, log checkpoints will never incur any 'collateral damage' of writing out unmodified changes for temporary tables. After this change, pages of the temporary tablespace can only be written out by buf_flush_lists(n_pages,0) as part of LRU eviction. Hopefully, most of the time, that code will never be executed, and instead, the temporary pages will be evicted by buf_release_freed_page() without ever being written back to the temporary tablespace file. This should improve the efficiency of the checkpoint flushing and the buf_flush_page_cleaner thread. Reviewed by: Vladislav Vaintroub
2020-12-09 08:22:13 +01:00
ut_ad(elem->oldest_modification());
ut_ad(!fsp_is_system_temporary(elem->id().space()));
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
}
};
MDEV-15058: Deprecate and ignore innodb_buffer_pool_instances Our benchmarking efforts indicate that the reasons for splitting the buf_pool in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 have mostly gone away, possibly as a result of mysql/mysql-server@ce6109ebfdedfdf185e391a0c97dc6d33867ed78 or similar work. Only in one write-heavy benchmark where the working set size is ten times the buffer pool size, the buf_pool->mutex would be less contended with 4 buffer pool instances than with 1 instance, in buf_page_io_complete(). That contention could be alleviated further by making more use of std::atomic and by splitting buf_pool_t::mutex further (MDEV-15053). We will deprecate and ignore the following parameters: innodb_buffer_pool_instances innodb_page_cleaners There will be only one buffer pool and one page cleaner task. In a number of INFORMATION_SCHEMA views, columns that indicated the buffer pool instance will be removed: information_schema.innodb_buffer_page.pool_id information_schema.innodb_buffer_page_lru.pool_id information_schema.innodb_buffer_pool_stats.pool_id information_schema.innodb_cmpmem.buffer_pool_instance information_schema.innodb_cmpmem_reset.buffer_pool_instance
2020-02-12 13:45:21 +01:00
/** Validate the flush list. */
static void buf_flush_validate_low()
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
{
buf_page_t* bpage;
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
mysql_mutex_assert_owner(&buf_pool.flush_list_mutex);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-21962 Allocate buf_pool statically Thanks to MDEV-15058, there is only one InnoDB buffer pool. Allocating buf_pool statically removes one level of pointer indirection and makes code more readable, and removes the awkward initialization of some buf_pool members. While doing this, we will also declare some buf_pool_t data members private and replace some functions with member functions. This is mostly affecting buffer pool resizing. This is not aiming to be a complete rewrite of buf_pool_t to a proper class. Most of the buffer pool interface, such as buf_page_get_gen(), will remain in the C programming style for now. buf_pool_t::withdrawing: Replaces buf_pool_withdrawing. buf_pool_t::withdraw_clock_: Replaces buf_withdraw_clock. buf_pool_t::create(): Repalces buf_pool_init(). buf_pool_t::close(): Replaces buf_pool_free(). buf_bool_t::will_be_withdrawn(): Replaces buf_block_will_be_withdrawn(), buf_frame_will_be_withdrawn(). buf_pool_t::clear_hash_index(): Replaces buf_pool_clear_hash_index(). buf_pool_t::get_n_pages(): Replaces buf_pool_get_n_pages(). buf_pool_t::validate(): Replaces buf_validate(). buf_pool_t::print(): Replaces buf_print(). buf_pool_t::block_from_ahi(): Replaces buf_block_from_ahi(). buf_pool_t::is_block_field(): Replaces buf_pointer_is_block_field(). buf_pool_t::is_block_mutex(): Replaces buf_pool_is_block_mutex(). buf_pool_t::is_block_lock(): Replaces buf_pool_is_block_lock(). buf_pool_t::is_obsolete(): Replaces buf_pool_is_obsolete(). buf_pool_t::io_buf: Make default-constructible. buf_pool_t::io_buf::create(): Delayed 'constructor' buf_pool_t::io_buf::close(): Early 'destructor' HazardPointer: Make default-constructible. Define all member functions inline, also for derived classes.
2020-03-18 20:48:00 +01:00
ut_list_validate(buf_pool.flush_list, Check());
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
MDEV-21962 Allocate buf_pool statically Thanks to MDEV-15058, there is only one InnoDB buffer pool. Allocating buf_pool statically removes one level of pointer indirection and makes code more readable, and removes the awkward initialization of some buf_pool members. While doing this, we will also declare some buf_pool_t data members private and replace some functions with member functions. This is mostly affecting buffer pool resizing. This is not aiming to be a complete rewrite of buf_pool_t to a proper class. Most of the buffer pool interface, such as buf_page_get_gen(), will remain in the C programming style for now. buf_pool_t::withdrawing: Replaces buf_pool_withdrawing. buf_pool_t::withdraw_clock_: Replaces buf_withdraw_clock. buf_pool_t::create(): Repalces buf_pool_init(). buf_pool_t::close(): Replaces buf_pool_free(). buf_bool_t::will_be_withdrawn(): Replaces buf_block_will_be_withdrawn(), buf_frame_will_be_withdrawn(). buf_pool_t::clear_hash_index(): Replaces buf_pool_clear_hash_index(). buf_pool_t::get_n_pages(): Replaces buf_pool_get_n_pages(). buf_pool_t::validate(): Replaces buf_validate(). buf_pool_t::print(): Replaces buf_print(). buf_pool_t::block_from_ahi(): Replaces buf_block_from_ahi(). buf_pool_t::is_block_field(): Replaces buf_pointer_is_block_field(). buf_pool_t::is_block_mutex(): Replaces buf_pool_is_block_mutex(). buf_pool_t::is_block_lock(): Replaces buf_pool_is_block_lock(). buf_pool_t::is_obsolete(): Replaces buf_pool_is_obsolete(). buf_pool_t::io_buf: Make default-constructible. buf_pool_t::io_buf::create(): Delayed 'constructor' buf_pool_t::io_buf::close(): Early 'destructor' HazardPointer: Make default-constructible. Define all member functions inline, also for derived classes.
2020-03-18 20:48:00 +01:00
bpage = UT_LIST_GET_FIRST(buf_pool.flush_list);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
while (bpage != NULL) {
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
const lsn_t om = bpage->oldest_modification();
MDEV-21962 Allocate buf_pool statically Thanks to MDEV-15058, there is only one InnoDB buffer pool. Allocating buf_pool statically removes one level of pointer indirection and makes code more readable, and removes the awkward initialization of some buf_pool members. While doing this, we will also declare some buf_pool_t data members private and replace some functions with member functions. This is mostly affecting buffer pool resizing. This is not aiming to be a complete rewrite of buf_pool_t to a proper class. Most of the buffer pool interface, such as buf_page_get_gen(), will remain in the C programming style for now. buf_pool_t::withdrawing: Replaces buf_pool_withdrawing. buf_pool_t::withdraw_clock_: Replaces buf_withdraw_clock. buf_pool_t::create(): Repalces buf_pool_init(). buf_pool_t::close(): Replaces buf_pool_free(). buf_bool_t::will_be_withdrawn(): Replaces buf_block_will_be_withdrawn(), buf_frame_will_be_withdrawn(). buf_pool_t::clear_hash_index(): Replaces buf_pool_clear_hash_index(). buf_pool_t::get_n_pages(): Replaces buf_pool_get_n_pages(). buf_pool_t::validate(): Replaces buf_validate(). buf_pool_t::print(): Replaces buf_print(). buf_pool_t::block_from_ahi(): Replaces buf_block_from_ahi(). buf_pool_t::is_block_field(): Replaces buf_pointer_is_block_field(). buf_pool_t::is_block_mutex(): Replaces buf_pool_is_block_mutex(). buf_pool_t::is_block_lock(): Replaces buf_pool_is_block_lock(). buf_pool_t::is_obsolete(): Replaces buf_pool_is_obsolete(). buf_pool_t::io_buf: Make default-constructible. buf_pool_t::io_buf::create(): Delayed 'constructor' buf_pool_t::io_buf::close(): Early 'destructor' HazardPointer: Make default-constructible. Define all member functions inline, also for derived classes.
2020-03-18 20:48:00 +01:00
/* A page in buf_pool.flush_list can be in
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
BUF_BLOCK_REMOVE_HASH state. This happens when a page
is in the middle of being relocated. In that case the
original descriptor can have this state and still be
in the flush list waiting to acquire the
MDEV-21962 Allocate buf_pool statically Thanks to MDEV-15058, there is only one InnoDB buffer pool. Allocating buf_pool statically removes one level of pointer indirection and makes code more readable, and removes the awkward initialization of some buf_pool members. While doing this, we will also declare some buf_pool_t data members private and replace some functions with member functions. This is mostly affecting buffer pool resizing. This is not aiming to be a complete rewrite of buf_pool_t to a proper class. Most of the buffer pool interface, such as buf_page_get_gen(), will remain in the C programming style for now. buf_pool_t::withdrawing: Replaces buf_pool_withdrawing. buf_pool_t::withdraw_clock_: Replaces buf_withdraw_clock. buf_pool_t::create(): Repalces buf_pool_init(). buf_pool_t::close(): Replaces buf_pool_free(). buf_bool_t::will_be_withdrawn(): Replaces buf_block_will_be_withdrawn(), buf_frame_will_be_withdrawn(). buf_pool_t::clear_hash_index(): Replaces buf_pool_clear_hash_index(). buf_pool_t::get_n_pages(): Replaces buf_pool_get_n_pages(). buf_pool_t::validate(): Replaces buf_validate(). buf_pool_t::print(): Replaces buf_print(). buf_pool_t::block_from_ahi(): Replaces buf_block_from_ahi(). buf_pool_t::is_block_field(): Replaces buf_pointer_is_block_field(). buf_pool_t::is_block_mutex(): Replaces buf_pool_is_block_mutex(). buf_pool_t::is_block_lock(): Replaces buf_pool_is_block_lock(). buf_pool_t::is_obsolete(): Replaces buf_pool_is_obsolete(). buf_pool_t::io_buf: Make default-constructible. buf_pool_t::io_buf::create(): Delayed 'constructor' buf_pool_t::io_buf::close(): Early 'destructor' HazardPointer: Make default-constructible. Define all member functions inline, also for derived classes.
2020-03-18 20:48:00 +01:00
buf_pool.flush_list_mutex to complete the relocation. */
MDEV-23399 fixup: Assertion bpage->in_file() failed buf_flush_remove_pages(), buf_flush_dirty_pages(): Because buf_page_t::state() is protected by buf_pool.mutex, which we are not holding, the state may be BUF_BLOCK_REMOVE_HASH when the page is being relocated. Let us relax these assertions similar to buf_flush_validate_low(). The other in_file() assertions in buf0flu.cc look valid.
2020-10-26 16:07:17 +01:00
ut_d(const auto s= bpage->state());
MDEV-27058: Reduce the size of buf_block_t and buf_page_t buf_page_t::frame: Moved from buf_block_t::frame. All 'thin' buf_page_t describing compressed-only ROW_FORMAT=COMPRESSED pages will have frame=nullptr, while all 'fat' buf_block_t will have a non-null frame pointing to aligned innodb_page_size bytes. This eliminates the need for separate states for BUF_BLOCK_FILE_PAGE and BUF_BLOCK_ZIP_PAGE. buf_page_t::lock: Moved from buf_block_t::lock. That is, all block descriptors will have a page latch. The IO_PIN state that was used for discarding or creating the uncompressed page frame of a ROW_FORMAT=COMPRESSED block is replaced by a combination of read-fix and page X-latch. page_zip_des_t::fix: Replaces state_, buf_fix_count_, io_fix_, status of buf_page_t with a single std::atomic<uint32_t>. All modifications will use store(), fetch_add(), fetch_sub(). This space was previously wasted to alignment on 64-bit systems. We will use the following encoding that combines a state (partly read-fix or write-fix) and a buffer-fix count: buf_page_t::NOT_USED=0 (previously BUF_BLOCK_NOT_USED) buf_page_t::MEMORY=1 (previously BUF_BLOCK_MEMORY) buf_page_t::REMOVE_HASH=2 (previously BUF_BLOCK_REMOVE_HASH) buf_page_t::FREED=3 + fix: pages marked as freed in the file buf_page_t::UNFIXED=1U<<29 + fix: normal pages buf_page_t::IBUF_EXIST=2U<<29 + fix: normal pages; may need ibuf merge buf_page_t::REINIT=3U<<29 + fix: reinitialized pages (skip doublewrite) buf_page_t::READ_FIX=4U<<29 + fix: read-fixed pages (also X-latched) buf_page_t::WRITE_FIX=5U<<29 + fix: write-fixed pages (also U-latched) buf_page_t::WRITE_FIX_IBUF=6U<<29 + fix: write-fixed; may have ibuf buf_page_t::WRITE_FIX_REINIT=7U<<29 + fix: write-fixed (no doublewrite) buf_page_t::write_complete(): Change WRITE_FIX or WRITE_FIX_REINIT to UNFIXED, and WRITE_FIX_IBUF to IBUF_EXIST, before releasing the U-latch. buf_page_t::read_complete(): Renamed from buf_page_read_complete(). Change READ_FIX to UNFIXED or IBUF_EXIST, before releasing the X-latch. buf_page_t::can_relocate(): If the page latch is being held or waited for, or the block is buffer-fixed or io-fixed, return false. (The condition on the page latch is new.) Outside buf_page_get_gen(), buf_page_get_low() and buf_page_free(), we will acquire the page latch before fix(), and unfix() before unlocking. buf_page_t::flush(): Replaces buf_flush_page(). Optimize the handling of FREED pages. buf_pool_t::release_freed_page(): Assume that buf_pool.mutex is held by the caller. buf_page_t::is_read_fixed(), buf_page_t::is_write_fixed(): New predicates. buf_page_get_low(): Ignore guesses that are read-fixed because they may not yet be registered in buf_pool.page_hash and buf_pool.LRU. buf_page_optimistic_get(): Acquire latch before buffer-fixing. buf_page_make_young(): Leave read-fixed blocks alone, because they might not be registered in buf_pool.LRU yet. recv_sys_t::recover_deferred(), recv_sys_t::recover_low(): Possibly fix MDEV-26326, by holding a page X-latch instead of only buffer-fixing the page.
2021-11-16 18:55:06 +01:00
ut_ad(s >= buf_page_t::REMOVE_HASH);
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
ut_ad(om == 1 || om > 2);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
bpage = UT_LIST_GET_NEXT(list, bpage);
MDEV-25113: Make page flushing faster buf_page_write_complete(): Reduce the buf_pool.mutex hold time, and do not acquire buf_pool.flush_list_mutex at all. Instead, mark blocks clean by setting oldest_modification to 1. Dirty pages of temporary tables will be identified by the special value 2 instead of the previous special value 1. (By design of the ib_logfile0 format, actual LSN values smaller than 2048 are not possible.) buf_LRU_free_page(), buf_pool_t::get_oldest_modification() and many other functions will remove the garbage (clean blocks) from buf_pool.flush_list while holding buf_pool.flush_list_mutex. buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaced with non-atomic variables, protected by buf_pool.mutex, to avoid unnecessary synchronization when modifying the counts. export_vars: Remove unnecessary indirection for innodb_pages_created, innodb_pages_read, innodb_pages_written.
2021-06-23 12:13:11 +02:00
ut_ad(om == 1 || !bpage || recv_recovery_is_on()
MDEV-23399: Remove buf_pool.flush_rbt Normally, buf_pool.flush_list must be sorted by buf_page_t::oldest_modification, so that log_checkpoint() can choose MIN(oldest_modification) as the checkpoint LSN. During recovery, buf_pool.flush_rbt used to guarantee the ordering. However, we can allow the buf_pool.flush_list to be in an arbitrary order during recovery, and simply ensure that it is in the correct order by the time a log checkpoint needs to be executed. recv_sys_t::apply(): To keep it simple, we will always flush the buffer pool at the end of each batch. Note that log_checkpoint() will invoke recv_sys_t::apply() in case a checkpoint is initiated during the last batch of recovery, when we already allow writes to data pages and the redo log. Reviewed by: Vladislav Vaintroub
2020-10-15 09:27:25 +02:00
|| om >= bpage->oldest_modification());
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
}
}
MDEV-15058: Deprecate and ignore innodb_buffer_pool_instances Our benchmarking efforts indicate that the reasons for splitting the buf_pool in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 have mostly gone away, possibly as a result of mysql/mysql-server@ce6109ebfdedfdf185e391a0c97dc6d33867ed78 or similar work. Only in one write-heavy benchmark where the working set size is ten times the buffer pool size, the buf_pool->mutex would be less contended with 4 buffer pool instances than with 1 instance, in buf_page_io_complete(). That contention could be alleviated further by making more use of std::atomic and by splitting buf_pool_t::mutex further (MDEV-15053). We will deprecate and ignore the following parameters: innodb_buffer_pool_instances innodb_page_cleaners There will be only one buffer pool and one page cleaner task. In a number of INFORMATION_SCHEMA views, columns that indicated the buffer pool instance will be removed: information_schema.innodb_buffer_page.pool_id information_schema.innodb_buffer_page_lru.pool_id information_schema.innodb_buffer_pool_stats.pool_id information_schema.innodb_cmpmem.buffer_pool_instance information_schema.innodb_cmpmem_reset.buffer_pool_instance
2020-02-12 13:45:21 +01:00
/** Validate the flush list. */
void buf_flush_validate()
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
{
MDEV-23399: Performance regression with write workloads The buffer pool refactoring in MDEV-15053 and MDEV-22871 shifted the performance bottleneck to the page flushing. The configuration parameters will be changed as follows: innodb_lru_flush_size=32 (new: how many pages to flush on LRU eviction) innodb_lru_scan_depth=1536 (old: 1024) innodb_max_dirty_pages_pct=90 (old: 75) innodb_max_dirty_pages_pct_lwm=75 (old: 0) Note: The parameter innodb_lru_scan_depth will only affect LRU eviction of buffer pool pages when a new page is being allocated. The page cleaner thread will no longer evict any pages. It used to guarantee that some pages will remain free in the buffer pool. Now, we perform that eviction 'on demand' in buf_LRU_get_free_block(). The parameter innodb_lru_scan_depth(srv_LRU_scan_depth) is used as follows: * When the buffer pool is being shrunk in buf_pool_t::withdraw_blocks() * As a buf_pool.free limit in buf_LRU_list_batch() for terminating the flushing that is initiated e.g., by buf_LRU_get_free_block() The parameter also used to serve as an initial limit for unzip_LRU eviction (evicting uncompressed page frames while retaining ROW_FORMAT=COMPRESSED pages), but now we will use a hard-coded limit of 100 or unlimited for invoking buf_LRU_scan_and_free_block(). The status variables will be changed as follows: innodb_buffer_pool_pages_flushed: This includes also the count of innodb_buffer_pool_pages_LRU_flushed and should work reliably, updated one by one in buf_flush_page() to give more real-time statistics. The function buf_flush_stats(), which we are removing, was not called in every code path. For both counters, we will use regular variables that are incremented in a critical section of buf_pool.mutex. Note that show_innodb_vars() directly links to the variables, and reads of the counters will *not* be protected by buf_pool.mutex, so you cannot get a consistent snapshot of both variables. The following INFORMATION_SCHEMA.INNODB_METRICS counters will be removed, because the page cleaner no longer deals with writing or evicting least recently used pages, and because the single-page writes have been removed: * buffer_LRU_batch_flush_avg_time_slot * buffer_LRU_batch_flush_avg_time_thread * buffer_LRU_batch_flush_avg_time_est * buffer_LRU_batch_flush_avg_pass * buffer_LRU_single_flush_scanned * buffer_LRU_single_flush_num_scan * buffer_LRU_single_flush_scanned_per_call When moving to a single buffer pool instance in MDEV-15058, we missed some opportunity to simplify the buf_flush_page_cleaner thread. It was unnecessarily using a mutex and some complex data structures, even though we always have a single page cleaner thread. Furthermore, the buf_flush_page_cleaner thread had separate 'recovery' and 'shutdown' modes where it was waiting to be triggered by some other thread, adding unnecessary latency and potential for hangs in relatively rarely executed startup or shutdown code. The page cleaner was also running two kinds of batches in an interleaved fashion: "LRU flush" (writing out some least recently used pages and evicting them on write completion) and the normal batches that aim to increase the MIN(oldest_modification) in the buffer pool, to help the log checkpoint advance. The buf_pool.flush_list flushing was being blocked by buf_block_t::lock for no good reason. Furthermore, if the FIL_PAGE_LSN of a page is ahead of log_sys.get_flushed_lsn(), that is, what has been persistently written to the redo log, we would trigger a log flush and then resume the page flushing. This would unnecessarily limit the performance of the page cleaner thread and trigger the infamous messages "InnoDB: page_cleaner: 1000ms intended loop took 4450ms. The settings might not be optimal" that were suppressed in commit d1ab89037a518fcffbc50c24e4bd94e4ec33aed0 unless log_warnings>2. Our revised algorithm will make log_sys.get_flushed_lsn() advance at the start of buf_flush_lists(), and then execute a 'best effort' to write out all pages. The flush batches will skip pages that were modified since the log was written, or are are currently exclusively locked. The MDEV-13670 message "page_cleaner: 1000ms intended loop took" message will be removed, because by design, the buf_flush_page_cleaner() should not be blocked during a batch for extended periods of time. We will remove the single-page flushing altogether. Related to this, the debug parameter innodb_doublewrite_batch_size will be removed, because all of the doublewrite buffer will be used for flushing batches. If a page needs to be evicted from the buffer pool and all 100 least recently used pages in the buffer pool have unflushed changes, buf_LRU_get_free_block() will execute buf_flush_lists() to write out and evict innodb_lru_flush_size pages. At most one thread will execute buf_flush_lists() in buf_LRU_get_free_block(); other threads will wait for that LRU flushing batch to finish. To improve concurrency, we will replace the InnoDB ib_mutex_t and os_event_t native mutexes and condition variables in this area of code. Most notably, this means that the buffer pool mutex (buf_pool.mutex) is no longer instrumented via any InnoDB interfaces. It will continue to be instrumented via PERFORMANCE_SCHEMA. For now, both buf_pool.flush_list_mutex and buf_pool.mutex will be declared with MY_MUTEX_INIT_FAST (PTHREAD_MUTEX_ADAPTIVE_NP). The critical sections of buf_pool.flush_list_mutex should be shorter than those for buf_pool.mutex, because in the worst case, they cover a linear scan of buf_pool.flush_list, while the worst case of a critical section of buf_pool.mutex covers a linear scan of the potentially much longer buf_pool.LRU list. mysql_mutex_is_owner(), safe_mutex_is_owner(): New predicate, usable with SAFE_MUTEX. Some InnoDB debug assertions need this predicate instead of mysql_mutex_assert_owner() or mysql_mutex_assert_not_owner(). buf_pool_t::n_flush_LRU, buf_pool_t::n_flush_list: Replaces buf_pool_t::init_flush[] and buf_pool_t::n_flush[]. The number of active flush operations. buf_pool_t::mutex, buf_pool_t::flush_list_mutex: Use mysql_mutex_t instead of ib_mutex_t, to have native mutexes with PERFORMANCE_SCHEMA and SAFE_MUTEX instrumentation. buf_pool_t::done_flush_LRU: Condition variable for !n_flush_LRU. buf_pool_t::done_flush_list: Condition variable for !n_flush_list. buf_pool_t::do_flush_list: Condition variable to wake up the buf_flush_page_cleaner when a log checkpoint needs to be written or the server is being shut down. Replaces buf_flush_event. We will keep using timed waits (the page cleaner thread will wake _at least_ once per second), because the calculations for innodb_adaptive_flushing depend on fixed time intervals. buf_dblwr: Allocate statically, and move all code to member functions. Use a native mutex and condition variable. Remove code to deal with single-page flushing. buf_dblwr_check_block(): Make the check debug-only. We were spending a significant amount of execution time in page_simple_validate_new(). flush_counters_t::unzip_LRU_evicted: Remove. IORequest: Make more members const. FIXME: m_fil_node should be removed. buf_flush_sync_lsn: Protect by std::atomic, not page_cleaner.mutex (which we are removing). page_cleaner_slot_t, page_cleaner_t: Remove many redundant members. pc_request_flush_slot(): Replaces pc_request() and pc_flush_slot(). recv_writer_thread: Remove. Recovery works just fine without it, if we simply invoke buf_flush_sync() at the end of each batch in recv_sys_t::apply(). recv_recovery_from_checkpoint_finish(): Remove. We can simply call recv_sys.debug_free() directly. srv_started_redo: Replaces srv_start_state. SRV_SHUTDOWN_FLUSH_PHASE: Remove. logs_empty_and_mark_files_at_shutdown() can communicate with the normal page cleaner loop via the new function flush_buffer_pool(). buf_flush_remove(): Assert that the calling thread is holding buf_pool.flush_list_mutex. This removes unnecessary mutex operations from buf_flush_remove_pages() and buf_flush_dirty_pages(), which replace buf_LRU_flush_or_remove_pages(). buf_flush_lists(): Renamed from buf_flush_batch(), with simplified interface. Return the number of flushed pages. Clarified comments and renamed min_n to max_n. Identify LRU batch by lsn=0. Merge all the functions buf_flush_start(), buf_flush_batch(), buf_flush_end() directly to this function, which was their only caller, and remove 2 unnecessary buf_pool.mutex release/re-acquisition that we used to perform around the buf_flush_batch() call. At the start, if not all log has been durably written, wait for a background task to do it, or start a new task to do it. This allows the log write to run concurrently with our page flushing batch. Any pages that were skipped due to too recent FIL_PAGE_LSN or due to them being latched by a writer should be flushed during the next batch, unless there are further modifications to those pages. It is possible that a page that we must flush due to small oldest_modification also carries a recent FIL_PAGE_LSN or is being constantly modified. In the worst case, all writers would then end up waiting in log_free_check() to allow the flushing and the checkpoint to complete. buf_do_flush_list_batch(): Clarify comments, and rename min_n to max_n. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_flush_space(): Auxiliary function to look up a tablespace for page flushing. buf_flush_page(): Defer the computation of space->full_crc32(). Never call log_write_up_to(), but instead skip persistent pages whose latest modification (FIL_PAGE_LSN) is newer than the redo log. Also skip pages on which we cannot acquire a shared latch without waiting. buf_flush_try_neighbors(): Do not bother checking buf_fix_count because buf_flush_page() will no longer wait for the page latch. Take the tablespace as a parameter, and only execute this function when innodb_flush_neighbors>0. Avoid repeated calls of page_id_t::fold(). buf_flush_relocate_on_flush_list(): Declare as cold, and push down a condition from the callers. buf_flush_check_neighbor(): Take id.fold() as a parameter. buf_flush_sync(): Ensure that the buf_pool.flush_list is empty, because the flushing batch will skip pages whose modifications have not yet been written to the log or were latched for modification. buf_free_from_unzip_LRU_list_batch(): Remove redundant local variables. buf_flush_LRU_list_batch(): Let the caller buf_do_LRU_batch() initialize the counters, and report n->evicted. Cache the last looked up tablespace. If neighbor flushing is not applicable, invoke buf_flush_page() directly, avoiding a page lookup in between. buf_do_LRU_batch(): Return the number of pages flushed. buf_LRU_free_page(): Only release and re-acquire buf_pool.mutex if adaptive hash index entries are pointing to the block. buf_LRU_get_free_block(): Do not wake up the page cleaner, because it will no longer perform any useful work for us, and we do not want it to compete for I/O while buf_flush_lists(innodb_lru_flush_size, 0) writes out and evicts at most innodb_lru_flush_size pages. (The function buf_do_LRU_batch() may complete after writing fewer pages if more than innodb_lru_scan_depth pages end up in buf_pool.free list.) Eliminate some mutex release-acquire cycles, and wait for the LRU flush batch to complete before rescanning. buf_LRU_check_size_of_non_data_objects(): Simplify the code. buf_page_write_complete(): Remove the parameter evict, and always evict pages that were part of an LRU flush. buf_page_create(): Take a pre-allocated page as a parameter. buf_pool_t::free_block(): Free a pre-allocated block. recv_sys_t::recover_low(), recv_sys_t::apply(): Preallocate the block while not holding recv_sys.mutex. During page allocation, we may initiate a page flush, which in turn may initiate a log flush, which would require acquiring log_sys.mutex, which should always be acquired before recv_sys.mutex in order to avoid deadlocks. Therefore, we must not be holding recv_sys.mutex while allocating a buffer pool block. BtrBulk::logFreeCheck(): Skip a redundant condition. row_undo_step(): Do not invoke srv_inc_activity_count() for every row that is being rolled back. It should suffice to invoke the function in trx_flush_log_if_needed() during trx_t::commit_in_memory() when the rollback completes. sync_check_enable(): Remove. We will enable innodb_sync_debug from the very beginning. Reviewed by: Vladislav Vaintroub
2020-10-15 11:10:42 +02:00
mysql_mutex_lock(&buf_pool.flush_list_mutex);
buf_flush_validate_low();
mysql_mutex_unlock(&buf_pool.flush_list_mutex);
initial commit into innodb-5.6 mergetree
2014-02-26 19:11:54 +01:00
}
MDEV-15053 Reduce buf_pool_t::mutex contention User-visible changes: The INFORMATION_SCHEMA views INNODB_BUFFER_PAGE and INNODB_BUFFER_PAGE_LRU will report a dummy value FLUSH_TYPE=0 and will no longer report the PAGE_STATE value READY_FOR_USE. We will remove some fields from buf_page_t and move much code to member functions of buf_pool_t and buf_page_t, so that the access rules of data members can be enforced consistently. Evicting or adding pages in buf_pool.LRU will remain covered by buf_pool.mutex. Evicting or adding pages in buf_pool.page_hash will remain covered by both buf_pool.mutex and the buf_pool.page_hash X-latch. After this fix, buf_pool.page_hash lookups can entirely avoid acquiring buf_pool.mutex, only relying on buf_pool.hash_lock_get() S-latch. Similarly, buf_flush_check_neighbors() can will rely solely on buf_pool.mutex, no buf_pool.page_hash latch at all. The buf_pool.mutex is rather contended in I/O heavy benchmarks, especially when the workload does not fit in the buffer pool. The first attempt to alleviate the contention was the buf_pool_t::mutex split in commit 4ed7082eefe56b3e97e0edefb3df76dd7ef5e858 which introduced buf_block_t::mutex, which we are now removing. Later, multiple instances of buf_pool_t were introduced in commit c18084f71b02ea707c6461353e6cfc15d7553bc6 and recently removed by us in commit 1a6f708ec594ac0ae2dd30db926ab07b100fa24b (MDEV-15058). UNIV_BUF_DEBUG: Remove. This option to enable some buffer pool related debugging in otherwise non-debug builds has not been used for years. Instead, we have been using UNIV_DEBUG, which is enabled in CMAKE_BUILD_TYPE=Debug. buf_block_t::mutex, buf_pool_t::zip_mutex: Remove. We can mainly rely on std::atomic and the buf_pool.page_hash latches, and in some cases depend on buf_pool.mutex or buf_pool.flush_list_mutex just like before. We must always release buf_block_t::lock before invoking unfix() or io_unfix(), to prevent a glitch where a block that was added to the buf_pool.free list would apper X-latched. See commit c5883debd6ef440a037011c11873b396923e93c5 how this glitch was finally caught in a debug environment. We move some buf_pool_t::page_hash specific code from the ha and hash modules to buf_pool, for improved readability. buf_pool_t::close(): Assert that all blocks are clean, except on aborted startup or crash-like shutdown. buf_pool_t::validate(): No longer attempt to validate n_flush[] against the number of BUF_IO_WRITE fixed blocks, because buf_page_t::flush_type no longer exists. buf_pool_t::watch_set(): Replaces buf_pool_watch_set(). Reduce mutex contention by separating the buf_pool.watch[] allocation and the insert into buf_pool.page_hash. buf_pool_t::page_hash_lock<bool exclusive>(): Acquire a buf_pool.page_hash latch. Replaces and extends buf_page_hash_lock_s_confirm() and buf_page_hash_lock_x_confirm(). buf_pool_t::READ_AHEAD_PAGES: Renamed from BUF_READ_AHEAD_PAGES. buf_pool_t::curr_size, old_size, read_ahead_area, n_pend_reads: Use Atomic_counter. buf_pool_t::running_out(): Replaces buf_LRU_buf_pool_running_out(). buf_pool_t::LRU_remove(): Remove a block from the LRU list and return its predecessor. Incorporates buf_LRU_adjust_hp(), which was removed. buf_page_get_gen(): Remove a redundant call of fsp_is_system_temporary(), for mode == BUF_GET_IF_IN_POOL_OR_WATCH, which is only used by BTR_DELETE_OP (purge), which is never invoked on temporary tables. buf_free_from_unzip_LRU_list_batch(): Avoid redundant assignments. buf_LRU_free_from_unzip_LRU_list(): Simplify the loop condition. buf_LRU_free_page(): Clarify the function comment. buf_flush_check_neighbor(), buf_flush_check_neighbors(): Rewrite the construction of the page hash range. We will hold the buf_pool.mutex for up to buf_pool.read_ahead_area (at most 64) consecutive lookups of buf_pool.page_hash. buf_flush_page_and_try_neighbors(): Remove. Merge to its only callers, and remove redundant operations in buf_flush_LRU_list_batch(). buf_read_ahead_random(), buf_read_ahead_linear(): Rewrite. Do not acquire buf_pool.mutex, and iterate directly with page_id_t. ut_2_power_up(): Remove. my_round_up_to_next_power() is inlined and avoids any loops. fil_page_get_prev(), fil_page_get_next(), fil_addr_is_null(): Remove. buf_flush_page(): Add a fil_space_t* parameter. Minimize the buf_pool.mutex hold time. buf_pool.n_flush[] is no longer updated atomically with the io_fix, and we will protect most buf_block_t fields with buf_block_t::lock. The function buf_flush_write_block_low() is removed and merged here. buf_page_init_for_read(): Use static linkage. Initialize the newly allocated block and acquire the exclusive buf_block_t::lock while not holding any mutex. IORequest::IORequest(): Remove the body. We only need to invoke set_punch_hole() in buf_flush_page() and nowhere else. buf_page_t::flush_type: Remove. Replaced by IORequest::flush_type. This field is only used during a fil_io() call. That function already takes IORequest as a parameter, so we had better introduce for the rarely changing field. buf_block_t::init(): Replaces buf_page_init(). buf_page_t::init(): Replaces buf_page_init_low(). buf_block_t::initialise(): Initialise many fields, but keep the buf_page_t::state(). Both buf_pool_t::validate() and buf_page_optimistic_get() requires that buf_page_t::in_file() be protected atomically with buf_page_t::in_page_hash and buf_page_t::in_LRU_list. buf_page_optimistic_get(): Now that buf_block_t::mutex no longer exists, we must check buf_page_t::io_fix() after acquiring the buf_pool.page_hash lock, to detect whether buf_page_init_for_read() has been initiated. We will also check the io_fix() before acquiring hash_lock in order to avoid unnecessary computation. The field buf_block_t::modify_clock (protected by buf_block_t::lock) allows buf_page_optimistic_get() to validate the block. buf_page_t::real_size: Remove. It was only used while flushing pages of page_compressed tables. buf_page_encrypt(): Add an output parameter that allows us ot eliminate buf_page_t::real_size. Replace a condition with debug assertion. buf_page_should_punch_hole(): Remove. buf_dblwr_t::add_to_batch(): Replaces buf_dblwr_add_to_batch(). Add the parameter size (to replace buf_page_t::real_size). buf_dblwr_t::write_single_page(): Replaces buf_dblwr_write_single_page(). Add the parameter size (to replace buf_page_t::real_size). fil_system_t::detach(): Replaces fil_space_detach(). Ensure that fil_validate() will not be violated even if fil_system.mutex is released and reacquired. fil_node_t::complete_io(): Renamed from fil_node_complete_io(). fil_node_t::close_to_free(): Replaces fil_node_close_to_free(). Avoid invoking fil_node_t::close() because fil_system.n_open has already been decremented in fil_space_t::detach(). BUF_BLOCK_READY_FOR_USE: Remove. Directly use BUF_BLOCK_MEMORY. BUF_BLOCK_ZIP_DIRTY: Remove. Directly use BUF_BLOCK_ZIP_PAGE, and distinguish dirty pages by buf_page_t::oldest_modification(). BUF_BLOCK_POOL_WATCH: Remove. Use BUF_BLOCK_NOT_USED instead. This state was only being used for buf_page_t that are in buf_pool.watch. buf_pool_t::watch[]: Remove pointer indirection. buf_page_t::in_flush_list: Remove. It was set if and only if buf_page_t::oldest_modification() is nonzero. buf_page_decrypt_after_read(), buf_corrupt_page_release(), buf_page_check_corrupt(): Change the const fil_space_t* parameter to const fil_node_t& so that we can report the correct file name. buf_page_monitor(): Declare as an ATTRIBUTE_COLD global function. buf_page_io_complete(): Split to buf_page_read_complete() and buf_page_write_complete(). buf_dblwr_t::in_use: Remove. buf_dblwr_t::buf_block_array: Add IORequest::flush_t. buf_dblwr_sync_datafiles(): Remove. It was a useless wrapper of os_aio_wait_until_no_pending_writes(). buf_flush_write_complete(): Declare static, not global. Add the parameter IORequest::flush_t. buf_flush_freed_page(): Simplify the code. recv_sys_t::flush_lru: Renamed from flush_type and changed to bool. fil_read(), fil_write(): Replaced with direct use of fil_io(). fil_buffering_disabled(): Remove. Check srv_file_flush_method directly. fil_mutex_enter_and_prepare_for_io(): Return the resolved fil_space_t* to avoid a duplicated lookup in the caller. fil_report_invalid_page_access(): Clean up the parameters. fil_io(): Return fil_io_t, which comprises fil_node_t and error code. Always invoke fil_space_t::acquire_for_io() and let either the sync=true caller or fil_aio_callback() invoke fil_space_t::release_for_io(). fil_aio_callback(): Rewrite to replace buf_page_io_complete(). fil_check_pending_operations(): Remove a parameter, and remove some redundant lookups. fil_node_close_to_free(): Wait for n_pending==0. Because we no longer do an extra lookup of the tablespace between fil_io() and the completion of the operation, we must give fil_node_t::complete_io() a chance to decrement the counter. fil_close_tablespace(): Remove unused parameter trx, and document that this is only invoked during the error handling of IMPORT TABLESPACE. row_import_discard_changes(): Merged with the only caller, row_import_cleanup(). Do not lock up the data dictionary while invoking fil_close_tablespace(). logs_empty_and_mark_files_at_shutdown(): Do not invoke fil_close_all_files(), to avoid a !needs_flush assertion failure on fil_node_t::close(). innodb_shutdown(): Invoke os_aio_free() before fil_close_all_files(). fil_close_all_files(): Invoke fil_flush_file_spaces() to ensure proper durability. thread_pool::unbind(): Fix a crash that would occur on Windows after srv_thread_pool->disable_aio() and os_file_close(). This fix was submitted by Vladislav Vaintroub. Thanks to Matthias Leich and Axel Schwenke for extensive testing, Vladislav Vaintroub for helpful comments, and Eugene Kosov for a review.
2020-06-05 11:35:46 +02:00
#endif /* UNIV_DEBUG */
Reference in a new issue Copy permalink