Re: Replacing brick in replicated volume withoutreducing redundancy?

[Strahil <hunter86_bg@xxxxxxxxx>]

Maybe something like:

gluster volume add-brick replica 4  arbiter 1 newbrick:/path-to-brick

Best Regards,
Strahil Nikolov

On Jan 7, 2020 14:22, Stefan Seidel <privat@xxxxxxxxxxxxxxxxx> wrote:

Doesn't seem to be working, at least not in a straightforward way:

$ gluster volume add-brick backupv01 host-06:/data/glusterfs/backupv01/brick06/gv0
volume add-brick: failed: Incorrect number of bricks supplied 1 with count 3

$ gluster volume add-brick backupv01 replica 3 arbiter 1 host-06:/data/glusterfs/backupv01/brick06/gv0
volume add-brick: failed: Operation failed

(Probably because replica 3 arbiter 1 is the same as replica 2 arbiter 1)

Stefan

---

From: "Stefan" <gluster@stefanseidel.info>
To: "Strahil Nikolov" <hunter86_bg@yahoo.com>
Cc: "gluster-users" <gluster-users@gluster.org>
Sent: Tuesday, 7 January, 2020 11:22:02
Subject: Re: [Gluster-users] Replacing brick in replicated volume without        reducing redundancy?

Thanks, that sounds interesting. From what I understand, arbiter setups only support multiple-of-2 data bricks, but maybe it's possible to replace the arbiter with a data brick, then replace the old data brick with an arbiter again.

Stefan

---

From: "Strahil Nikolov" <hunter86_bg@yahoo.com>
To: "gluster-users" <gluster-users@gluster.org>, "Stefan" <gluster@stefanseidel.info>
Sent: Tuesday, 7 January, 2020 10:18:11
Subject: Re: [Gluster-users] Replacing brick in replicated volume without        reducing redundancy?

I'm not sure if you can increase (and later decrease) the replica count without stopping the volume.

Have you checked that as an option ?

Best Regards,

Strahil Nikolov

В вторник, 7 януари 2020 г., 3:58:11 ч. Гринуич-5, Stefan <gluster@stefanseidel.info> написа:

Hi,

I am looking to replace a brick in a replicated volume with arbiter (2+1 bricks).

Judging from this page: https://docs.gluster.org/en/latest/Administrator%20Guide/Managing%20Volumes/#replace-brick

it seems that the procedure is: simulate a brick failure, then add a replacement brick.

Is that the only way to do it? Isn't there a way to migrate the data on the fly without running on reduced redundancy during the migration, like pvmove exists for LVM?

Thanks,

Stefan

________

Community Meeting Calendar:

APAC Schedule -

________

Community Meeting Calendar:

APAC Schedule -
Every 2nd and 4th Tuesday at 11:30 AM IST
Bridge: https://bluejeans.com/441850968

NA/EMEA Schedule -
Every 1st and 3rd Tuesday at 01:00 PM EDT
Bridge: https://bluejeans.com/441850968

Gluster-users mailing list
Gluster-users@xxxxxxxxxxx
https://lists.gluster.org/mailman/listinfo/gluster-users