Comment 4 for bug 1516933

Ramesh Sivaraman (rameshvs02) wrote :

Used latest PXC version

mysql> show status like 'wsrep_provider_version';
+------------------------+------------------+
| Variable_name | Value |
+------------------------+------------------+
| wsrep_provider_version | 3.12.2(rf3e626d) |
+------------------------+------------------+
1 row in set (0.00 sec)

mysql> select @@version,@@version_comment;
+--------------------+---------------------------------------------------------------------------------------------+
| @@version | @@version_comment |
+--------------------+---------------------------------------------------------------------------------------------+
| 5.6.26-74.0-56-log | Percona XtraDB Cluster (GPL), Release rel74.0, Revision 1, WSREP version 25.12, wsrep_25.12 |
+--------------------+---------------------------------------------------------------------------------------------+
1 row in set (0.00 sec)

mysql>

I tried following steps to reproduce the issue

1) Download latest binary tar ball from https://www.percona.com/downloads/Percona-XtraDB-Cluster-56/LATEST/
2) Start PXC nodes with innodb_print_all_deadlocks=1
3) Simulate following deadlock tc to reproduce the issue

CREATE TABLE t1 (i int);
CREATE TABLE t2 (i int);

INSERT t1 SELECT 1;
INSERT t2 SELECT 9;

#session1
START TRANSACTION;
UPDATE t1 SET i=11 WHERE i=1;
select SLEEP(15);
UPDATE t2 SET i=99 WHERE i=9;
COMMIT;

#session 2
START TRANSACTION;
UPDATE t2 SET i=99 WHERE i=9;
select SLEEP(15);
UPDATE t1 SET i=11 WHERE i=1;
COMMIT;

4) Check deadlock status and error log

Can you re-check (probably with the steps I mentioned to be doubly sure)