GlusterFS - Out of memory - Killed

Issue

• We are experiencing some problems with Red Hat Storage. We have a volume from the RHS nodes mounted on a RHEL 6.4 client running the following version of glusterfs:

[root@server ~]# glusterfs --version
glusterfs 3.4.0.14rhs built on Jul 30 2013 09:19:58

• It works well for a limit period of time before glusterfs is killed with the following error:

Sep 10 20:33:05 server kernel: Out of memory: Kill process 3199 (glusterfs) score 960 or sacrifice child
Sep 10 20:33:05 server kernel: Killed process 3199, UID 0, (glusterfs) total-vm:265925140kB, anon-rss:261523348kB, file-rss:988kB

glusterfs invoked oom-killer while geo-replication on master node. :

[root@server ~]# glusterfs --version
glusterfs-3.4.0.59rhs built on Jul 30 2013 09:19:58

Jul 24 23:52:54 server kernel: Out of memory: Kill process 16132 (glusterfs) score 861 or sacrifice child
Jul 24 23:52:54 server kernel: Killed process 16132, UID 0, (glusterfs) total-vm:66538020kB, anon-rss:58844248kB, file-rss:2140kB