Gluster users - Dec 2017 - Bricks not starting after restart node

Hello Gluster Community,

I am using Gluster3.10 in a 2-Node-Szenario (with CentosOS7-machines).
After restarting one of the nodes,
sometimes one of these two phenomena occurs:

    1. The brick process of a volume on the newly started node is no
longer running
    2. The Brick Process Is Running But The First Knot Still Believes
He Can not Reach
    The Brick (Health info reported "Transport endpoint not
connected")

In both cases it helps to stop and restart the volume (gluster volume
stop, gluster volume start).
Sometimes you even have to do this twice! But this should be no
soluton in a productive scenario.
For ealier releases there was a bug posted here:
https://bugzilla.redhat.com/show_bug.cgi?id=1333711
It seems to be that this bug is still an issue? Does anybody has some
experience with that issue?

Regards

David Spisla
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.gluster.org/pipermail/gluster-users/attachments/20171220/9148ec7e/attachment.html>

Could you share the glusterd and the respective brick log files along with
the output of gluster volume info, gluster volume status and 'ps aux | grep
glusterfsd' .

On Wed, Dec 20, 2017 at 3:08 PM, David Spisla <spisla80 at googlemail.com>
wrote:
> Hello Gluster Community,
>
> I am using Gluster3.10 in a 2-Node-Szenario (with CentosOS7-machines).
After restarting one of the nodes,
> sometimes one of these two phenomena occurs:
>
>     1. The brick process of a volume on the newly started node is no longer
running
>     2. The Brick Process Is Running But The First Knot Still Believes He
Can not Reach
>     The Brick (Health info reported "Transport endpoint not
connected")
>
> In both cases it helps to stop and restart the volume (gluster volume stop,
gluster volume start).
> Sometimes you even have to do this twice! But this should be no soluton in
a productive scenario.
> For ealier releases there was a bug posted here:
https://bugzilla.redhat.com/show_bug.cgi?id=1333711
> It seems to be that this bug is still an issue? Does anybody has some
experience with that issue?
>
> Regards
>
> David Spisla
>
>
>
>
> _______________________________________________
> Gluster-users mailing list
> Gluster-users at gluster.org
> http://lists.gluster.org/mailman/listinfo/gluster-users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.gluster.org/pipermail/gluster-users/attachments/20171220/16e12601/attachment.html>