CRASHES INNODB | TRX_STATE_NOT_STARTED
The problem was that if DELETE with subselect caused a
deadlock inside InnoDB, this deadlock was not properly
handled by the SQL layer. This meant that the SQL layer
would try to unlock the row after InnoDB had rolled
back the transaction. This caused an assertion inside
InnoDB.
This patch fixes the problem by checking for errors
reported by SQL_SELECT::skip_record() and not calling
unlock_row() if any errors have been reported.
Problem
========
Replication breaks in the cases if the event length exceeds
the size of master Dump thread's max_allowed_packet.
The reason why this failure is occuring is because the event length is
more than the total size of the max_allowed_packet, on addition of the
max_event_header length exceeds the max_allowed_packet of the DUMP thread.
This causes the Dump thread to break replication and throw an error.
That can happen e.g with row-based replication in Update_rows event.
Fix
====
The problem is fixed in 2 steps:
1.) The Dump thread limit to read event is increased to the upper limit
i.e. Dump thread reads whatever gets logged in the binary log.
2.) On the slave side we increase the the max_allowed_packet for the
slave's threads (IO/SQL) by increasing it to 1GB.
This is done using the new server option (slave_max_allowed_packet)
included, is used to regulate the max_allowed_packet of the
slave thread (IO/SQL) by the DBA, and facilitates the sending of
large packets from the master to the slave.
This causes the large packets to be received by the slave and apply
it successfully.
sql/log_event.cc:
The max_allowed_packet is not evaluated to the new option
slave_max_allowed_packet after the fix.
sql/log_event.h:
Added the new option in the log_event.h file.
sql/mysqld.cc:
Added a new option to the server.
sql/slave.cc:
Increasing the session max_allowed_packet to a large value,
i.e. not taking global(max_allowed) into consideration, for the slave's threads.
sql/sql_repl.cc:
The dump thread's max_allowed_packet is set to the upper limit
which makes it independent and it now reads whatever gets
logged in the binary log.
One of the reported problems manifested itself in the scenario when one
thread tried to to get statistics on a key cache while the second thread
had not finished initialization of the key cache structure yet.
The problem was resolved by forcing serialization of such operations
on key caches.
To serialize function calls to perform certain operations over a key cache
a new mutex associated with the key cache now is used. It is stored in the
field op_lock of the KEY_CACHE structure. It is locked when the operation
is performed. Some of the serialized key cache operations utilize calls
for other key cache operations. To avoid recursive locking of op_lock
the new functions that perform the operations of key cache initialization,
destruction and re-partitioning with an additional parameter were introduced.
The parameter says whether the operation over op_lock are to be performed or
are to be omitted. The old functions for the operations of key cache
initialization, destruction,and re-partitioning now just call the
corresponding new functions with the additional parameter set to true
requesting to use op_lock while all other calls of these new function
have this parameter set to false.
Another problem reported in the bug entry concerned the operation of
assigning an index to a key cache. This operation can be called
while the key cache structures are not initialized yet. In this
case any call of flush_key_blocks() should return without any actions.
No test case is provided with this patch.
The bug prevented acceptance of UNION queries whose non-first select
clauses contained join expressions with degenerated single-table nests
as valid queries.
The bug was introduced into mysql-5.5 code line by the patch for
bug 33204.
Fixed MDEV-331: last_insert_id() returns a signed number
mysql-test/r/auto_increment.result:
Added test case
mysql-test/t/auto_increment.test:
Added test case
sql/item_func.h:
Changed last_insert_id() to be unsigned.
- Item::get_seconds() now skips decimal arithmetic, if decimals is 0. This significantly speeds up from_unixtime() if no fractional part is passed.
- replace sprintfs used to format temporal values by hand-coded formatting
Query1 (original query in the bug report)
BENCHMARK(10000000,DATE_SUB(FROM_UNIXTIME(RAND() * 2147483648), INTERVAL (FLOOR(1 + RAND() * 365)) DAY))
Query2 (Variation of query1 that does not use fractional part in FROM_UNIXTIME parameter)
BENCHMARK(10000000,DATE_SUB(FROM_UNIXTIME(FLOOR(RAND() * 2147483648)), INTERVAL (FLOOR(1 + RAND() * 365)) DAY))
Prior to the patch, the runtimes were (32 bit compilation/AMD machine)
Query1: 41.53 sec
Query2: 23.90 sec
With the patch, the runtimes are
Query1: 32.32 sec (speed up due to removing sprintf)
Query2: 12.06 sec (speed up due to skipping decimal arithmetic)
The --debug-no-sync incorrectly defaulted to ON, disabling sync calls
by default which can loose data or cause corruption. Also, the code
used fsync() instead of the sometimes more efficient fdatasync().
Changes in the InnoDB codebase required to compile and
integrate the MEB codebase with MySQL 5.5.
@ storage/innobase/btr/btr0btr.c
Excluded buffer pool usage from MEB build.
buf_pool_from_bpage calls are in buf0buf.ic, and
the buffer pool functions from that file are
disabled in MEB.
@ storage/innobase/buf/buf0buf.c
Disabling more buffer pool functions unused in MEB.
@ storage/innobase/dict/dict0dict.c
Disabling dict_ind_free that is unused in MEB.
@ storage/innobase/dict/dict0mem.c
The include
#include "ha_prototypes.h"
Was causing conflicts with definitions in my_global.h
Linking C executable mysqlbackup
libinnodb.a(dict0mem.c.o): In function `dict_mem_foreign_table_name_lookup_set':
dict0mem.c:(.text+0x91c): undefined reference to `innobase_get_lower_case_table_names'
libinnodb.a(dict0mem.c.o): In function `dict_mem_referenced_table_name_lookup_set':
dict0mem.c:(.text+0x9fc): undefined reference to `innobase_get_lower_case_table_names'
libinnodb.a(dict0mem.c.o): In function `dict_mem_foreign_table_name_lookup_set':
dict0mem.c:(.text+0x96e): undefined reference to `innobase_casedn_str'
libinnodb.a(dict0mem.c.o): In function `dict_mem_referenced_table_name_lookup_set':
dict0mem.c:(.text+0xa4e): undefined reference to `innobase_casedn_str'
collect2: ld returned 1 exit status
make[2]: *** [mysqlbackup] Error 1
innobase_get_lower_case_table_names
innobase_casedn_str
are functions that are part of ha_innodb.cc that is not part of the build
dict_mem_foreign_table_name_lookup_set
function is not there in the current codebase, meaning we do not use it in MEB.
@ storage/innobase/fil/fil0fil.c
The srv_fast_shutdown variable is declared in
srv0srv.c that is not compiled in the
mysqlbackup codebase.
This throws an undeclared error.
From the Manual
---------------
innodb_fast_shutdown
--------------------
The InnoDB shutdown mode. The default value is 1
as of MySQL 3.23.50, which causes a “fast� shutdown
(the normal type of shutdown). If the value is 0,
InnoDB does a full purge and an insert buffer merge
before a shutdown. These operations can take minutes,
or even hours in extreme cases. If the value is 1,
InnoDB skips these operations at shutdown.
This ideally does not matter from mysqlbackup
@ storage/innobase/ha/ha0ha.c
In file included from /home/narayanan/mysql-server/mysql-5.5-meb-rel3.8-innodb-integration-1/storage/innobase/ha/ha0ha.c:34:0:
/home/narayanan/mysql-server/mysql-5.5-meb-rel3.8-innodb-integration-1/storage/innobase/include/btr0sea.h:286:17: error: expected ‘=’, ‘,’, ‘;’, ‘asm’ or ‘__attribute__’ before ‘*’ token
make[2]: *** [CMakeFiles/innodb.dir/home/narayanan/mysql-server/mysql-5.5-meb-rel3.8-innodb-integration-1/storage/innobase/ha/ha0ha.c.o] Error 1
make[1]: *** [CMakeFiles/innodb.dir/all] Error 2
make: *** [all] Error 2
# include "sync0rw.h" is excluded from hotbackup compilation in dict0dict.h
This causes extern rw_lock_t* btr_search_latch_temp; to throw a failure because
the definition of rw_lock_t is not found.
@ storage/innobase/include/buf0buf.h
Excluding buffer pool functions that are unused from the
MEB codebase.
@ storage/innobase/include/buf0buf.ic
replicated the exclusion of
#include "buf0flu.h"
#include "buf0lru.h"
#include "buf0rea.h"
by looking at the current codebase in <meb-trunk>/src/innodb
@ storage/innobase/include/dict0dict.h
dict_table_x_lock_indexes, dict_table_x_unlock_indexes, dict_table_is_corrupted,
dict_index_is_corrupted, buf_block_buf_fix_inc_func are unused in MEB and was
leading to compilation errors and hence excluded.
@ storage/innobase/include/dict0dict.ic
dict_table_x_lock_indexes, dict_table_x_unlock_indexes, dict_table_is_corrupted,
dict_index_is_corrupted, buf_block_buf_fix_inc_func are unused in MEB and was
leading to compilation errors and hence excluded.
@ storage/innobase/include/log0log.h
/home/narayanan/mysql-server/mysql-5.5-meb-rel3.8-innodb-integration-1/storage/innobase/include/log0log.h: At top level:
/home/narayanan/mysql-server/mysql-5.5-meb-rel3.8-innodb-integration-1/storage/innobase/include/log0log.h:767:2: error: expected specifier-qualifier-list before â⠂¬Ëœmutex_t’
mutex_t definitions were excluded as seen from ambient code
hence excluding definition for log_flush_order_mutex also.
@ storage/innobase/include/os0file.h
Bug in InnoDB code, create_mode should have been create.
@ storage/innobase/include/srv0srv.h
In file included from /home/narayanan/mysql-server/mysql-5.5-meb-rel3.8-innodb-integration-1/storage/innobase/buf/buf0buf.c:50:0:
/home/narayanan/mysql-server/mysql-5.5-meb-rel3.8-innodb-integration-1/storage/innobase/include/srv0srv.h: At top level:
/home/narayanan/mysql-server/mysql-5.5-meb-rel3.8-innodb-integration-1/storage/innobase/include/srv0srv.h:120:16: error: expected ‘=’, ‘,’, ‘;’, ‘asm’ or ‘__attribute__’ before ‘srv_use_native_aio’
srv_use_native_aio - we do not use native aio of the OS anyway from MEB. MEB does not compile
InnoDB with this option. Hence disabling it.
@ storage/innobase/include/trx0sys.h
[ 56%] Building C object CMakeFiles/innodb.dir/home/narayanan/mysql-server/mysql-5.5-meb-rel3.8-innodb-integration-1/storage/innobase/trx/trx0sys.c.o
/home/narayanan/mysql-server/mysql-5.5-meb-rel3.8-innodb-integration-1/storage/innobase/trx/trx0sys.c: In function ‘trx_sys_read_file_format_id’:
/home/narayanan/mysql-server/mysql-5.5-meb-rel3.8-innodb-integration-1/storage/innobase/trx/trx0sys.c:1499:20: error: ‘TRX_SYS_FILE_FORMAT_TAG_MAGIC_N’ undeclared (first use in this function)
/home/narayanan/mysql-server/mysql-5.5-meb-rel3.8-innodb-integration-1/storage/innobase/trx/trx0sys.c:1499:20: note: each undeclared identifier is reported only once for each function it appears in
/home/narayanan/mysql-server/mysql-5.5-meb-rel3.8-innodb-integration-1/storage/innobase/trx/trx0sys.c: At top level:
/home/narayanan/mysql-server/mysql-5.5-meb-rel3.8-innodb-integration-1/storage/innobase/include/buf0buf.h:607:1: warning: ‘buf_block_buf_fix_inc_func’ declared ‘static’ but never defined
make[2]: *** [CMakeFiles/innodb.dir/home/narayanan/mysql-server/mysql-5.5-meb-rel3.8-innodb-integration-1/storage/innobase/trx/trx0sys.c.o] Error 1
unused calls excluded to enable compilation
@ storage/innobase/mem/mem0dbg.c
excluding #include "ha_prototypes.h" that lead to definitions in ha_innodb.cc
@ storage/innobase/os/os0file.c
InnoDB not compiled with aio support from MEB anyway. Hence excluding this from
the compilation.
@ storage/innobase/page/page0zip.c
page0zip.c:(.text+0x4e9e): undefined reference to `buf_pool_from_block'
collect2: ld returned 1 exit status
buf_pool_from_block defined in buf0buf.ic, most of the file is excluded for compilation of MEB
@ storage/innobase/ut/ut0dbg.c
excluding #include "ha_prototypes.h" since it leads to definitions in ha_innodb.cc
innobase_basename(file) is defined in ha_innodb.cc. Hence excluding that also.
@ storage/innobase/ut/ut0ut.c
cal_tm unused from MEB, was leading to earnings, hence disabling for MEB.
Analysis:
When the method JOIN::choose_subquery_plan() decided to apply
the IN-TO-EXISTS strategy, it set the unit and select_lex
uncacheable flag to UNCACHEABLE_DEPENDENT_INJECTED unconditionally.
As result, even if IN-TO-EXISTS injected non-correlated predicates,
the subquery was still treated as correlated.
Solution:
Set the subquery as correlated only if the injected predicate(s) depend
on the outer query.
backport dmitry.shulga@oracle.com-20120209125742-w7hdxv0103ymb8ko from mysql-trunk:
Patch for bug#11764747 (formerly known as 57612): SET GLOBAL READ_ONLY=1 cannot
progress when a table is locked with LOCK TABLES.
The reason for the bug was that mysql server makes a flush of all open tables
during handling of statement 'SET GLOBAL READ_ONLY=1'. Therefore if some of
these tables were locked by "LOCK TABLE ... READ" from a different connection,
then execution of statement 'SET GLOBAL READ_ONLY=1' would be waiting for
the lock for such table even if the table was locked in a compatible read mode.
Flushing of all open tables before setting of read_only system variable
is inherited from 5.1 implementation since this was the only possible approach
to ensure that there isn't any pending write operations on open tables.
Start from version 5.5 and above such behaviour is guaranteed by the fact
that we acquire global_read_lock before setting read_only flag. Since
acquiring of global_read_lock is successful only when there isn't any
active write operation then we can remove flushing of open tables from
processing of SET GLOBAL READ_ONLY=1.
This modification changes the server behavior so that read locks held
by other connections (LOCK TABLE ... READ) no longer will block attempts
to enable read_only.
Analysis:
The crash is a result of Item_cache_temporal::example not being set
(it is NULL). It turns out that the value of Item_cache_temporal
may be set directly by calling Item_cache_temporal::store_packed
without ever setting the "example" of this Item_cache. Therefore
the failing assertion is too narrow.
Solution:
Remove the assert.
In principle we could overwrite this method for Item_cache_temporal,
but it doesn't make sense just for this assert.
The constructor for Query_log_event allocated 2 bytes too few for
extra space needed by Query cache. (Not sure if this is reproducible
in practice, as there are often a couple of extra bytes allocated
for unused string zero terminators, but better safe than sorry).
CHEAP SQ: Valgrind warnings "Memory lost" with IN and EXISTS nested subquery, materialization+semijoin
Analysis:
The memory leak was a result of the interaction of semi-join optimization
with early optimization of constant subqueries. The function:
setup_jtbm_semi_joins() created a dummy temporary table "dummy_table"
in order to make some JOIN_TAB objects complete. Normally, such temporary
tables are freed inside JOIN_TAB::cleanup.
However, the inner-most subquery is pre-optimized, which allows the
optimization fo the MAX subquery to determine that its WHERE is TRUE,
and thus to compute the result of the MAX during optimization. This
ultimately allows the optimize phase of the outer query to find that
it WHERE clause is FALSE. Once JOIN::optimize finds that the result
set is empty, it sets zero_result_cause, and returns *before* it ever
reached make_join_statistics(). As a result the query plan has no
JOIN_TABs at all. Since the temporary table is supposed to be cleanup
via JOIN_TAB::cleanup, this never happens because there is no JOIN_TAB
for this table. Hence we get a memory leak.
Solution:
Whenever there are no JOIN_TABs, iterate over all table reference in
JOIN::join_list, and free the ones that contain semi-join temporary
tables.
WHEN KILLING
Suppose there is a query waiting for a lock. If the user kills
this query, then "Got error -1 when reading table" error message
must not be logged in the server log file. Since this is a user
requested interruption, no spurious error message must be logged
in the server log. This patch will remove the error message from
the log.
approved by joh and tatjana
Fixed by backport of:
------------------------------------------------------------
revno: 3402.50.156
committer: Jon Olav Hauglid <jon.hauglid@oracle.com>
branch nick: mysql-trunk-test
timestamp: Wed 2012-02-08 14:10:23 +0100
message:
Bug#13417754 ASSERT IN ROW_DROP_DATABASE_FOR_MYSQL DURING DROP SCHEMA
This assert could be triggered if an InnoDB table was being moved
to a different database using ALTER TABLE ... RENAME, while this
database concurrently was being dropped by DROP DATABASE.
The reason for the problem was that no metadata lock was taken
on the target database by ALTER TABLE ... RENAME.
DROP DATABASE was therefore not blocked and could remove
the database while ALTER TABLE ... RENAME was executing. This
could cause the assert in InnoDB to be triggered.
This patch fixes the problem by taking a IX metadata lock on
the target database before ALTER TABLE ... RENAME starts
moving a table to a different database.
Note that this problem did not occur with RENAME TABLE which
already takes the correct metadata locks.
Also note that this patch slightly changes the behavior of
ALTER TABLE ... RENAME. Before, the statement would abort and
return an error if a lock on the target table name could not
be taken immediately. With this patch, ALTER TABLE ... RENAME
will instead block and wait until the lock can be taken
(or until we get a lock timeout). This also means that it is
possible to get ER_LOCK_DEADLOCK errors in this situation
since we allow ALTER TABLE ... RENAME to wait and not just
abort immediately.
- Fixed code that was not ready for a major version number > 9
- Fixed test cases that assumed max major version number could be 9
Updated version number for depricated options (will be removed in a later commit)
VERSION:
Version number 10.0.0
client/mysqlbinlog.cc:
Added support for major version numbers > 9
cmake/mysql_version.cmake:
Added support for version numbers that is 0
mysql-test/r/comments.result:
Modified test to handle version number 100000
mysql-test/r/func_system.result:
Modified test to handle version number 100000
mysql-test/r/log_state.result:
Updated depricated error message
mysql-test/r/sp.result:
Modified test to handle version number 100000
mysql-test/r/subselect4.result:
Updated depricated error message
mysql-test/r/variables.result:
Updated depricated error message
mysql-test/suite/rpl/r/rpl_conditional_comments.result:
Modified test to handle version number 100000
mysql-test/suite/rpl/r/rpl_loaddatalocal.result:
Modified test to handle version number 100000
mysql-test/suite/rpl/t/rpl_conditional_comments.test:
Modified test to handle version number 100000
mysql-test/suite/rpl/t/rpl_loaddatalocal.test:
Modified test to handle version number 100000
mysql-test/suite/sys_vars/r/debug_basic.result:
Updated depricated error message
mysql-test/suite/sys_vars/r/engine_condition_pushdown_basic.result:
Updated depricated error message
mysql-test/suite/sys_vars/r/log_basic.result:
Updated depricated error message
mysql-test/suite/sys_vars/r/log_slow_queries_basic.result:
Updated depricated error message
mysql-test/suite/sys_vars/r/multi_range_count_basic.result:
Updated depricated error message
mysql-test/suite/sys_vars/r/rpl_recovery_rank_basic.result:
Updated depricated error message
mysql-test/suite/sys_vars/r/sql_big_selects_func.result:
Updated depricated error message
mysql-test/suite/sys_vars/r/sql_max_join_size_basic.result:
Updated depricated error message
mysql-test/suite/sys_vars/r/sql_max_join_size_func.result:
Updated depricated error message
mysql-test/t/comments.test:
Modified test to handle version number 100000
mysql-test/t/file_contents.test:
Modified test to handle version number 100000
mysql-test/t/func_system.test:
Modified test to handle version number 100000
mysql-test/t/parser_not_embedded.test:
Modified test to handle version number 100000
mysql-test/t/sp.test:
Modified test to handle version number 100000
sql/mysqld.cc:
Updated version number for depricated options (will be removed in a later commit)
sql/slave.cc:
Modified test to handle version number 100000
Better error messages
sql/sql_lex.cc:
Modified test to handle version number 100000 in comment syntax
sql/sys_vars.cc:
Updated version number for depricated options (will be removed in a later commit)
