database/mariadb
1
0
Fork 0
mirror of https://github.com/MariaDB/server.git synced 2025-07-27 18:02:13 +03:00
Code Activity

MDEV-8113: Parallel slave: slave hangs on ALTER TABLE (or other DDL) as the first event after slave start

Browse Source 
In optimistic parallel replication, it is not safe to try to run a following
transaction in parallel with a DDL statement, and there is code to prevent
this.

However, the code was missing the case where the DDL is the very first event
after slave start. In this case, following transactions could run in
parallel with the DDL, which can cause the slave to hang or even corrupt
slave in unlucky cases.
This commit is contained in:
Kristian Nielsen
2015-05-11 11:55:58 +02:00
parent ecfc3de57e
commit 8bedb638d7
3 changed files with 73 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

41
mysql-test/suite/rpl/r/rpl_parallel_optimistic.result
View File

@ -455,6 +455,47 @@ a b
include/stop_slave.inc
SET GLOBAL debug_dbug= @old_debug;
include/start_slave.inc
*** MDEV-8113: ALTER TABLE causes slave hang in optimistic parallel replication ***
include/stop_slave.inc
ALTER TABLE t2 ADD c INT;
INSERT INTO t2 (a,b) VALUES (50, 0);
INSERT INTO t2 (a,b) VALUES (51, 1);
INSERT INTO t2 (a,b) VALUES (52, 2);
INSERT INTO t2 (a,b) VALUES (53, 3);
INSERT INTO t2 (a,b) VALUES (54, 4);
INSERT INTO t2 (a,b) VALUES (55, 5);
INSERT INTO t2 (a,b) VALUES (56, 6);
INSERT INTO t2 (a,b) VALUES (57, 7);
INSERT INTO t2 (a,b) VALUES (58, 8);
INSERT INTO t2 (a,b) VALUES (59, 9);
ALTER TABLE t2 DROP COLUMN c;
SELECT * FROM t2 WHERE a >= 50 ORDER BY a;
a b
50 0
51 1
52 2
53 3
54 4
55 5
56 6
57 7
58 8
59 9
include/save_master_gtid.inc
include/start_slave.inc
include/sync_with_master_gtid.inc
SELECT * FROM t2 WHERE a >= 50 ORDER BY a;
a b
50 0
51 1
52 2
53 3
54 4
55 5
56 6
57 7
58 8
59 9
include/stop_slave.inc
SET GLOBAL slave_parallel_mode=@old_parallel_mode;
SET GLOBAL slave_parallel_threads=@old_parallel_threads;