when replicate-wild-do-table result in OOM

[ypli]

slave : 5.5.29-29.4-log Percona Server (GPL), Release rel29.4, Revision 401
master：Server version: 5.1.47-rel11.1-log Percona Server (GPL), 11.1, Revision 51

slave memory 72G buffer pool 45G ,no other services（only replicat 3 table use replicate-wild-do-table）

and the slave‘s memory utilize get more and more ,about 2G add every day, until no swap no memory , then OOM ,mysql restart

the dmesg info about mysql

Out of memory: Kill process 16856 (mysqld) score 985 or sacrifice child
Killed process 16856, UID 497, (mysqld) total-vm:78087080kB, anon-rss:72909800kB, file-rss:2696kB

and I remove the replicate-wild-do-table， add slave-skip-errors=1146 in my.cnf， and restart mysql, everything right,

the memory utilize is 70.8% of 72G total, no further incresed

is this a bug about the slave sql thread？

[Valerii Kravchuk (valerii-kravchuk)]

Looks like a duplicate of upstream http://bugs.mysql.com/bug.php?id=67983 and/or Percoan Server bug https://bugs.launchpad.net/percona-server/+bug/1042946 (that is already fixed).