simple working solution for such cases is rebuilding volume with .glusterfs
recover dead node, create fresh bricks, copy files there and then attr them
to regenerate .glusterfs
something like mentioned here
https://lists.gluster.org/pipermail/gluster-users/2018-January/033352.html
On Mon, Oct 29, 2018 at 7:46 AM Geoffray Levasseur <
fatalerrors at geoffray-levasseur.org> wrote:
> Hi everyone,
>
> I have a serious issue with one of my peer on a mirrored nodes with an
> arbiter. Both nodes
> operates a ZFS filesystem where GlusterFS is on top of it.
>
> Because of a malfunctioning controller one of the ZFS filesystem tell me
> my entire pool is corrupted and I need to destroy then recreate the entire
> pool. Anyway I want to use that to reshape the ZFS filesystem so it's
what
> I'll do.
>
> Can you tell me if a documentation exists for such situation (no server
> change, no ID change, etc.) so I can restore all the data after ZFS
> filesystem recreation?
>
> Thank you very much,
> --
> Geoffray Levasseur
> Technicien UPS - UMR CNRS 5566 / LEGOS - Service Informatique
> <fatalerrors at geoffray-levasseur.org>
> <geoffray.levasseur at legos.obs-mip.fr>
> http://www.geoffray-levasseur.org
> Quod gratis asseritur gratis negatur.
> _______________________________________________
> Gluster-users mailing list
> Gluster-users at gluster.org
> https://lists.gluster.org/mailman/listinfo/gluster-users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.gluster.org/pipermail/gluster-users/attachments/20181031/2c6572e2/attachment.html>