On Tue, Feb 21, 2017 at 6:15 PM, Jeff Darcy <jdarcy@xxxxxxxxxx> wrote:
> > 2) Bug 1421590 - Bricks take up new ports upon volume restart after
> > add-brick op with brick mux enabled
> > - Status: *Atin/Samikshan/Jeff*, any update on this?
> > - Can we document this as a known issue? What would be the way to
> > get volume to use the older ports (a glusterd restart?)?
That would work, but is clearly less than ideal.
>
> Patch under review for master. Atin/Samikshan are we going to wait for
> this be backported? The last update was we need not consider this a
> blocker for the release, does that still hold?
I can review/backport this today, despite being on vacation (we're all
giving our legs a day off). Does that help?
I don't think its a blocker. So irrespective of whether the patch gets in or not, the release should not be blocked. We can mark it as a known issue in the release note.
> > 4) Bug 1422769 - brick process crashes when glusterd is restarted
>
> Atin/Samikshan, thoughts on this?
Very slight possibility this was an after-effect of 1421721, which is
fixed. Hard to tell, though, since I was never able to reproduce it
on my systems.
Again, not a blocker and I was unable to hit it with multiple attempts.
> > 5) Bug 1422781 - Transport endpoint not connected error seen on client
> > when glusterd is restarted
> > - Status: Repro not clean across setups, still debugging the problem
>
> Atin/Samikshan, were we able to narrow this down, post attempts to
> reproduce it?
Still trying to figure out why this won't reproduce using the cluster.rc
stuff. There seemed to be some possibility that it was related to the
amount of I/O that was active while GlusterD was restarted, or to things
being in containers, but haven't heard back from Atin.
Although I was able to hit this issue at first attempt on release-3.10 head when the bug was filed, but later I couldn't with multiple attempts. Not sure any patch went in between which has fixed it. Given its not consistent, probably we can live with it.
However I am a bit bothered about https://bugzilla.redhat.com/show_bug.cgi?id=1421724 where we see a flood of log entries in glusterd.log if a volume is restarted after turning on brick multiplexing and this scenario does look to be hit in the set up frequently. I did work with Samikshan & Gaurav and figured out a possible RCA. I have posted a patch https://review.gluster.org/#/c/16699, reviews please?
_______________________________________________
Gluster-devel mailing list
Gluster-devel@xxxxxxxxxxx
http://lists.gluster.org/mailman/listinfo/gluster-devel
--
~ Atin (atinm)
_______________________________________________ Gluster-devel mailing list Gluster-devel@xxxxxxxxxxx http://lists.gluster.org/mailman/listinfo/gluster-devel
