The problem is the async binlog checkpointing; this could on rare
occasions occur too late, causing SHOW BINLOG EVENTS to show the
wrong events and cause .result file difference.
Replication caches the character sets used in a query, to be able to quickly
reuse them for the next query in the common case of them not having changed.
In parallel replication, this caching needs to be per-worker-thread. The
code was not modified to handle this correctly, so the caching in one worker
could cause another worker to run a query using the wrong character set,
causing replication corruption.
WRITTEN WHILE ROWS REMAINS
Problem:
========
When truncate table fails while using transactional based
engines even though the operation errors out we still
continue and log it to binlog. Because of this master has
data but the truncate will be written to binary log which
will cause inconsistency.
Analysis:
========
Truncate table can happen either through drop and create of
table or by deleting rows. In the second case the existing
code is written in such a way that even if an error occurs
the truncate statement will always be binlogged. Which is not
correct.
Binlogging of TRUNCATE TABLE statement should check whether
truncate is executed "transactionally or not". If the table
is transaction based we log the TRUNCATE TABLE only on
successful completion.
If table is non transactional there are possibilities that on
error we could have partial changes done hence in such cases
we do log in spite of errors as some of the lines might have
been removed, so the statement has to be sent to slave.
Fix:
===
Using table handler whether truncate table is being executed
in transaction based mode or not is identified and statement
is binlogged accordingly.
mysql-test/suite/binlog/r/binlog_truncate_kill.result:
Added test case to test the fix for Bug#17942050.
mysql-test/suite/binlog/t/binlog_truncate_kill.test:
Added test case to test the fix for Bug#17942050.
sql/sql_truncate.cc:
Check if truncation is successful or not and retun appropriate
return values so that binlogging can be done based on that.
sql/sql_truncate.h:
Added a new enum.
This will help the following replicaition scenario:
MariaDB 10.0 master (statement replication) -> MariaDB 10.0 slave (row based replication) -> MySQL or MariaDB 5.x slave
mysql-test/r/mysqld--help.result:
Updated help text
mysql-test/suite/rpl/r/create_or_replace_mix.result:
Added more tests
mysql-test/suite/rpl/r/create_or_replace_row.result:
Added more tests
mysql-test/suite/rpl/r/create_or_replace_statement.result:
Added more tests
mysql-test/suite/rpl/t/create_or_replace.inc:
Added more tests
sql/handler.h:
Added org_options so that we can detect what come from the query and what was possible added later.
sql/sql_insert.cc:
Only write CREATE OR REPLACE if was originally specified or if we delete a conflicting table as part of create
sql/sql_parse.cc:
Remember orginal create options
sql/sql_table.cc:
Only write CREATE OR REPLACE if was originally specified or if we delete a conflicting table as part of create
sql/sys_vars.cc:
Updated help text
Copied relevant test cases and code from the MySQL 5.6 tree
Testing of my_use_symdir moved to engines.
mysql-test/r/partition_windows.result:
Updated result file
mysql-test/suite/archive/archive_no_symlink-master.opt:
Testing of symlinks with archive
mysql-test/suite/archive/archive_no_symlink.result:
Testing of symlinks with archive
mysql-test/suite/archive/archive_no_symlink.test:
Testing of symlinks with archive
mysql-test/suite/archive/archive_symlink.result:
Testing of symlinks with archive
mysql-test/suite/archive/archive_symlink.test:
Testing of symlinks with archive
sql/log_event.cc:
Updated comment
sql/partition_info.cc:
Don't test my_use_symdir here
sql/sql_parse.cc:
Updated comment
sql/sql_table.cc:
Don't test my_use_symdir here
sql/table.cc:
Added more DBUG_PRINT
storage/archive/ha_archive.cc:
Give warnings for index_file_name and if we can't use data directory
storage/myisam/ha_myisam.cc:
Give warnings if we can't use data directory or index directory
Bug #3329 Incomplete lower_case_table_names=2 implementation
The problem was that check_db_name() converted database names to lower case also in case of lower_case_table_names=2.
Fixed by removing the conversion in check_db_name for lower_case_table_names = 2 and instead converting db name to
lower case at same places as table names are converted.
Fixed bug that SHOW CREATE DATABASE FOO showed information for database 'foo'.
I also removed some checks of lower_case_table_names when it was enough to use table_alias_charset.
mysql-test/mysql-test-run.pl:
Added --use-copy argument to force mysql-test-run to copy files instead of doing symlinks. This is needed when you run
with test directory on another file system
mysql-test/r/lowercase_table.result:
Updated results
mysql-test/r/lowercase_table2.result:
Updated results
mysql-test/suite/parts/r/partition_mgm_lc2_innodb.result:
Updated results
mysql-test/suite/parts/r/partition_mgm_lc2_memory.result:
Updated results
mysql-test/suite/parts/r/partition_mgm_lc2_myisam.result:
Updated results
mysql-test/t/lowercase_table.test:
Added tests with mixed case databases
mysql-test/t/lowercase_table2.test:
Added tests with mixed case databases
sql/log.cc:
Don't check lower_case_table_names when we can use table_alias_charset
sql/sql_base.cc:
Don't check lower_case_table_names when we can use table_alias_charset
sql/sql_db.cc:
Use cmp_db_names() for checking if current database changed.
mysql_rm_db() now converts db to lower case if lower_case_table_names was used.
Changed database options cache to use table_alias_charset. This fixed a bug where SHOW CREATE DATABASE showed wrong information.
sql/sql_parse.cc:
Change also db name to lower case when file names are changed.
Don't need to story copy of database name anymore when lower_case_table_names == 2 as check_db_name() don't convert in this case.
Updated arguments to mysqld_show_create_db().
When adding table to TABLE_LIST also convert db name to lower case if needed (same way as we do with table names).
sql/sql_show.cc:
mysqld_show_create_db() now also takes original name as argument for output to user.
sql/sql_show.h:
Updated prototype for mysqld_show_create_db()
sql/sql_table.cc:
In mysql_rename_table(), do same conversions to database name as we do for the file name
The reason was that a couple of variables that hold number of rows that was used to calculate buffers was uint and caused an overflow.
Fixed by changing variables that could hold number of rows from uint to ulong and also added a cast for this test.
include/heap.h:
Reorder to get better alignment. Changed variables that could hold number of rows from uint to ulong
mysql-test/suite/heap/heap.result:
Added test case
mysql-test/suite/heap/heap.test:
Added test case
mysql-test/suite/plugins/t/server_audit.test:
Added sleep as we want to have disconnect logged before we try a new connect
storage/heap/ha_heap.cc:
Changed variables that could hold number of rows from uint to ulong
Limit number of rows to 4G (as most of the variables that holds rows are ulong anyway)
reset records_changed when key_stat_version is changed to not cause increments for every row changed
storage/heap/ha_heap.h:
changed records_changed to ulong as this can get big
storage/heap/hp_create.c:
Changed variables that could hold number of rows from uint to ulong
Added cast (fixed the original bug)
storage/heap/hp_delete.c:
Changed variables that could hold number of rows from uint to ulong
storage/heap/hp_open.c:
Removed not needed cast
storage/heap/hp_write.c:
Changed variables that could hold number of rows from uint to ulong
support-files/compiler_warnings.supp:
Removed extra : from supression
Due to how gap locks work, two transactions could group commit together on the
master, but get lock conflicts and then deadlock due to different thread
scheduling order on slave.
For now, remove these deadlocks by running the parallel slave in READ
COMMITTED mode. And let InnoDB/XtraDB allow statement-based binlogging for the
parallel slave in READ COMMITTED.
We are also investigating a different solution long-term, which is based on
relaxing the gap locks only between the transactions running in parallel for
one slave, but not against possibly external transactions.
When a transaction fails in parallel replication, it should signal the error
to any following transactions doing wait_for_prior_commit() on it. But the
code for this was incorrect, and would not correctly remember a prior error
when sending the signal. This caused corruption when slave stopped due to an
error.
Fix by remembering the error code when we first get an error, and passing the
saved error code to wakeup_subsequent_commits().
Thanks to nanyi607rao who reported this bug on
maria-developers@lists.launchpad.net and analysed the root cause.
Let TABLE_SHARE::tdc.free_tables, TABLE_SHARE::tdc.all_tables,
TABLE_SHARE::tdc.flushed and corresponding invariants be protected by
per-share TABLE_SHARE::tdc.LOCK_table_share instead of global LOCK_open.
Now if CREATE OR REPLACE fails but we have deleted a table already, we will generate a DROP TABLE in the binary log.
This fixes this issue.
In addition, for a failing CREATE OR REPLACE TABLE ... SELECT we don't generate a log of all the inserted rows, only the DROP TABLE.
I added code for not logging DROP TEMPORARY TABLE for tables where the CREATE TABLE was not logged. This code will be activated in 10.1
by removing the code protected by DONT_LOG_DROP_OF_TEMPORARY_TABLES.
mysql-test/suite/rpl/r/create_or_replace_mix.result:
More test cases
mysql-test/suite/rpl/r/create_or_replace_row.result:
More test cases
mysql-test/suite/rpl/r/create_or_replace_statement.result:
More test cases
mysql-test/suite/rpl/t/create_or_replace.inc:
More test cases
sql/log.cc:
Added binlog_reset_cache() to clear the binary log.
sql/log.h:
Added prototype
sql/sql_insert.cc:
If CREATE OR REPLACE TABLE ... SELECT fails:
- Don't log anything if nothing changed
- If table was deleted, log a DROP TABLE.
Remember if we table creation of temporary tables was logged.
sql/sql_table.cc:
Added log_drop_table()
Remember if we table creation of temporary tables was logged.
If CREATE OR REPLACE TABLE ... SELECT fails and a table was deleted, log a DROP TABLE.
sql/sql_table.h:
Added prototype
sql/sql_truncate.cc:
Remember if we table creation of temporary tables was logged.
sql/table.h:
Added table_creation_was_logged
mysql-test/suite/rpl/t/rpl_000011-slave.opt:
Renamed test case as it's slave that needs to restarted
support-files/compiler_warnings.supp:
Fixed bad characters in suppression
mysql-test/suite/rpl/t/rpl_000011-master.opt:
Added master.opt file to ensure that other tests don't interfere with rpl_000011
plugin/server_audit/server_audit.c:
Fixed compiler error on solaris
support-files/compiler_warnings.supp:
Ignore warning from xtradb
Reason for the bug was an optimization for higher connect speed where we moved when global status was updated,
but forgot to update states when slave thread dies.
Fixed by adding thd->add_status_to_global() before deleting slave thread's thd.
mysys/my_delete.c:
Added missing newline
sql/mysqld.cc:
Use add_status_to_global()
sql/slave.cc:
Added missing add_status_to_global()
sql/sql_class.cc:
Use add_status_to_global()
sql/sql_class.h:
Simplify adding local status to global by adding add_status_to_global()
Before, the arrival of same GTID twice in multi-source replication
would cause double-apply or in gtid strict mode an error.
Keep the behaviour, but add an option --gtid-ignore-duplicates which
allows to correctly handle duplicates, ignoring all but the first.
This relies on the user ensuring correct configuration so that
sequence numbers are strictly increasing within each replication
domain; then duplicates can be detected simply by comparing the
sequence numbers against what is already applied.
Only one master connection (but possibly multiple parallel worker
threads within that connection) is allowed to apply events within
one replication domain at a time; any other connection that
receives a GTID in the same domain either discards it (if it is
already applied) or waits for the other connection to not have
any events to apply.
Intermediate patch, as proof-of-concept for testing. The main limitation
is that currently it is only implemented for parallel replication,
@@slave_parallel_threads > 0.
Automatic merge, except for server_audit.cc that had to be modified slightly
Changes to xtradb and innobase where ignored was these made no sence for 10.0
When an rpl_group_info object was returned from the free list, the
rgi->deferred_events_collecting and rgi->deferred_events was not correctly
re-inited. Additionally, the rgi->deferred_events was incorrectly freed in
free_rgi(), which causes unnecessary malloc/free (or crash when re-init is not
done).
Thanks to user nanyi607rao, who reported this bug on maria-developers@.
