Convoy-gluster file changed as we read it

[najunju111]

I have 3 hosts.installed gluster-fs and convoy-gluster from catalog. I modified my_vol types, from reply to Distribute (/opt/rancher/replicated_volume_create.sh num 58: gluster volume create $ {VOLUME_NAME} transport tcp $ {CONTAINER_MNTS}) I created a volume from UI named pools_1 and mount it to a container(volumes is pools_1:/tmp1). but docker-enter in container,cd /tmp1.When I run the command: tar zcvf openvpn-2.3.10.tar.gz ./openvpn-2.3.10, display an error: file changed as we read it If my_vol mount the hard disk directory, without Convoy-gluster, there is no mistake

[pierreozoux]

Ok, I hit the same bug!

(it is also mentionned in the forum here)

Looks like the solution is:

upgrade to 3.7.12:
You have to have the following configuration:
1) gluster volume set <volname> cluster.consistent-metadata on
2) gluster volume set <volname> cluster.post-op-delay-secs 0

from: https://bugzilla.redhat.com/show_bug.cgi?id=1058526

Would you have some indication on how I can help on that?

[yasker]

Sorry we're stopping support for glusterfs now due to resource limitation... Put this in low priority.