database/mariadb
1
0
Fork 0
mirror of https://github.com/MariaDB/server.git synced 2025-07-29 05:21:33 +03:00
Code Activity

MDEV-6551: Some replication errors are ignored if slave_parallel_threads > 0

Browse Source 
The problem occured when using parallel replication, and an error occured that
caused the SQL thread to stop when the IO thread had already reached a
following binlog file from the master (or otherwise performed a relay log
rotation).

In this case, the Rotate Event at the end of the relay log file could still be
executed, even though an earlier event in that relay log file had gotten an
error. This would cause the position to be incorrectly updated, so that upon
restart of the SQL thread, the event that had failed would be silently skipped
and ignored, causing replication corruption.

Fixed by checking before executing Rotate Event, whether an earlier event
has failed. If so, the Rotate Event is not executed, just dequeued, same as
for other normal events following a failing event.
This commit is contained in:
Kristian Nielsen
2014-08-15 11:31:13 +02:00
parent 65ac881c80
commit cfa1ce81bb
4 changed files with 131 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files

58
mysql-test/suite/rpl/t/rpl_parallel.test
View File

@ -1408,6 +1408,64 @@ SELECT * FROM t6 ORDER BY a;
SELECT * FROM t6 ORDER BY a;
--echo *** MDEV-6551: Some replication errors are ignored if slave_parallel_threads > 0 ***
--connection server_1
INSERT INTO t2 VALUES (31);
--source include/save_master_gtid.inc
--connection server_2
--source include/sync_with_master_gtid.inc
--source include/stop_slave.inc
SET GLOBAL slave_parallel_threads= 0;
--source include/start_slave.inc
# Force a duplicate key error on the slave.
SET sql_log_bin= 0;
INSERT INTO t2 VALUES (32);
SET sql_log_bin= 1;
--connection server_1
INSERT INTO t2 VALUES (32);
# Rotate the binlog; the bug is triggered when the master binlog file changes
# after the event group that causes the duplicate key error.
FLUSH LOGS;
INSERT INTO t2 VALUES (33);
INSERT INTO t2 VALUES (34);
SELECT * FROM t2 WHERE a >= 30 ORDER BY a;
--source include/save_master_gtid.inc
--connection server_2
--let $slave_sql_errno= 1062
--source include/wait_for_slave_sql_error.inc
--connection server_2
--source include/stop_slave_io.inc
SET GLOBAL slave_parallel_threads=10;
START SLAVE;
--let $slave_sql_errno= 1062
--source include/wait_for_slave_sql_error.inc
# Note: IO thread is still running at this point.
# The bug seems to have been that restarting the SQL thread after an error with
# the IO thread still running, somehow picks up a later relay log position and
# thus ends up skipping the failing event, rather than re-executing.
START SLAVE SQL_THREAD;
--let $slave_sql_errno= 1062
--source include/wait_for_slave_sql_error.inc
SELECT * FROM t2 WHERE a >= 30 ORDER BY a;
# Skip the duplicate error, so we can proceed.
SET sql_slave_skip_counter= 1;
--source include/start_slave.inc
--source include/sync_with_master_gtid.inc
SELECT * FROM t2 WHERE a >= 30 ORDER BY a;
--connection server_2
--source include/stop_slave.inc
SET GLOBAL slave_parallel_threads=@old_parallel_threads;