We had a distributed replicated volume of 3 x 7 HDD, the volume was used for small files workload with heavy IO, we decided to replace the bricks with SSDs because of IO saturation to the disks, so we started by swapping the bricks one by one, and the fun started, some files lost its attributes and we had to manually fix the missing attributes by removing the file and its gfid and copy the file again to the volume.
This issue affected 5 of the 21 bricks.
On another volume, we had a disk failure and during the replace brick process, the mount point of one of the clients crashed.
On Mon, Jun 22, 2020 at 10:55 AM Gionatan Danti <g.danti@xxxxxxxxxx> wrote:
Il 2020-06-21 20:41 Mahdi Adnan ha scritto:
> Hello Gionatan,
>
> Using Gluster brick in a RAID configuration might be safer and
> require less work from Gluster admins but, it is a waste of disk
> space.
> Gluster bricks are replicated "assuming you're creating a
> distributed-replica volume" so when brick went down, it should be easy
> to recover it and should not affect the client's IO.
> We are using JBOD in all of our Gluster setups, overall, performance
> is good, and replacing a brick would work "most" of the time without
> issues.
Hi Mahdi,
thank you for reporting. I am interested in the "most of the time
without isses" statement. Can you elaborate on what happened the few
times when it did not work correctly?
Thanks.
--
Danti Gionatan
Supporto Tecnico
Assyoma S.r.l. - www.assyoma.it [1]
email: g.danti@xxxxxxxxxx - info@xxxxxxxxxx
GPG public key ID: FF5F32A8
Respectfully
Mahdi
________ Community Meeting Calendar: Schedule - Every 2nd and 4th Tuesday at 14:30 IST / 09:00 UTC Bridge: https://bluejeans.com/441850968 Gluster-users mailing list Gluster-users@xxxxxxxxxxx https://lists.gluster.org/mailman/listinfo/gluster-users