But if you *do* decide to test the feature, I'd suggest that you do so with 3.8.2 (which is slated to be released on Aug 10) as it contains some important bug fixes. :)
On Sun, Aug 7, 2016 at 8:39 AM, David Gossage <dgossage@xxxxxxxxxxxxxxxxxx> wrote:
On Sat, Aug 6, 2016 at 9:58 PM, Lindsay Mathieson <lindsay.mathieson@xxxxxxxxx> wrote:On 7/08/2016 12:56 PM, David Gossage wrote:
My Dev server with one VM doing almost nothing handled update to 3.8.1 and op-version update just fine.
Good to know, thanks :)
turned on the new granular setting but haven't simulated any brick failures yet to see what happens.
is that the "cluster.locking-scheme: granular" one? its available in 3.7.12 as well, I've had it activate for a couple of weeks.Granular Entry self-heal for AFR
Notes for users: This feature can be enabled using the command
gluster volume set <vol-name> granular-entry-heal onLimitations: 1. The feature is not backward compatible. So please enable the option only after you have upgraded all your clients and servers to 3.8 and op-version is 30800 2. Make sure the volume is stopped and there is no pending heal before you enable the feature.Not even sure if I can test what it's meant to improve on my lackluster dev setup.
--
Lindsay Mathieson
_______________________________________________
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
