Re: Replica 3 cluster, file being healed on all 3 nodes

[Pranith Kumar Karampuri <pkarampu@xxxxxxxxxx>]

On 10/14/2015 09:39 AM, Lindsay Mathieson wrote:

On 13 October 2015 at 22:33, Krutika Dhananjay <kdhananj@xxxxxxxxxx> wrote:

However I managed to create a state where a file was being healed on all three nodes (probably y live migrating a VM while it was being healed). I didn't think that was possible without creating a split brain problem, but it eventually got all the way to being healed.

I don't think it is possible for heal of this image to be happening on all three nodes.

I should have recorded the info output, but it did show the same file being "possibly healed" on all three nodes.

There seems to be a confusion in understanding how to interpret output of "gluster volume heal <volname> info". We will address it by either improving the output or adding some documentation about how to interpret it.
For now, all it means is that one of the self-heal-daemons or the mount is doing the heal on that file. The main confusing point seems to be the fact that the same output is seen on multiple bricks. Finding intersection/union of the results between bricks will cost (LOT)more iops, so we went with giving same output multiple times(when same info is present on more than one brick) as perceived by each brick. Doesn't mean that each of the bricks is doing the heal!, afr takes necessary locks to make sure parallel heals don't happen on the file.

Pranith

Gluster 3.6.6

--

Lindsay

_______________________________________________
Gluster-users mailing list
Gluster-users@xxxxxxxxxxx
http://www.gluster.org/mailman/listinfo/gluster-users

_______________________________________________
Gluster-users mailing list
Gluster-users@xxxxxxxxxxx
http://www.gluster.org/mailman/listinfo/gluster-users