Hello,
--
I am running a single 3-node replicated cluster utilizing Glusterfs 3.6.1 on CentOS 6.6 with all updates as of this posting. The only 3 modification I have made are as follows:
gluster volume set gv0 cluster.data-self-heal off
gluster volume set gv0 cluster.entry-self-heal off
gluster volume set gv0 cluster.metadata-self-heal off
I just did a quick reboot of one node in the cluster and I was surprised to see the load spike for a few minutes and I found several entries such as follows in my glustershd.log:
[2015-01-13 22:26:58.649673] I [afr-self-heal-common.c:476:afr_log_selfheal] 0-gv0-replicate-0: Completed data selfheal on 4eafb61e-954f-4ee4-bc1e-e68fd1df5954. source=0 sinks=
[2015-01-13 22:29:01.443858] I [afr-self-heal-common.c:476:afr_log_selfheal] 0-gv0-replicate-0: Completed data selfheal on 338bfe7b-5cf7-407a-ae21-383e0d259ec7. source=2 sinks=0 1
[2015-01-13 22:30:03.298315] I [afr-self-heal-common.c:476:afr_log_selfheal] 0-gv0-replicate-0: Completed data selfheal on e91f0283-fe13-40c8-8c7c-6f93d8c54268. source=0 sinks=1
[2015-01-13 22:30:03.373751] I [afr-self-heal-metadata.c:41:__afr_selfheal_metadata_do] 0-gv0-replicate-0: performing metadata selfheal on e91f0283-fe13-40c8-8c7c-6f93d8c54268
[2015-01-13 22:30:03.483401] I [afr-self-heal-common.c:476:afr_log_selfheal] 0-gv0-replicate-0: Completed metadata selfheal on e91f0283-fe13-40c8-8c7c-6f93d8c54268. source=0 sinks=1
Given the settings I have above that should turn off metadata and data self heal, why is this occurring? How can I truly turn these off without turning the cluster.self-heal-daemon off? I would rather not do that as with it turned off I can't see what files need to be healed such that I can heal them during non-production hours.
Thank you.
Kyle
_______________________________________________ Gluster-users mailing list Gluster-users@xxxxxxxxxxx http://www.gluster.org/mailman/listinfo/gluster-users
