[Gluster-users] Urgent :) Procedure for replacing Gluster Node on 3.8.12

Pranith Kumar Karampuri pkarampu at redhat.com
Sat Jun 10 04:06:04 UTC 2017


On Sat, Jun 10, 2017 at 2:53 AM, Lindsay Mathieson <
lindsay.mathieson at gmail.com> wrote:

> On 9/06/2017 9:56 PM, Pranith Kumar Karampuri wrote:
>
> > gluster volume remove-brick datastore4 replica 2
>> > vna.proxmox.softlog:/tank/vmdata/datastore4 force
>> >
>> > gluster volume add-brick datastore4 replica 3
>> > vnd.proxmox.softlog:/tank/vmdata/datastore4
>>
>> I think that should work perfectly fine yes, either that
>> or directly use replace-brick ?
>>
>
> Yes, this should be replace-brick
>
>
> Was there any problem in doing the way I did?
>
You won't notice the problem. But it is inefficient.
AFR keeps track of which brick is good/bad based on the xattrs on the
bricks. If you have 3 nodes A,B,C then on each brick there can be at least
2-3 extra AFR xattrs of the form (trusted.afr.<volname>-client-[0/1/2].
When you replace C->D then the same xattrs will stay. But when you remove C
and add D it will treat it as a new brick and add one more xattr like
trusted.afr.<volname>-client-3. So it is better to use replace-brick for
this case.

Add-brick/remove-brick is for users who make the decision to make the
decision that they either want to increase the replica count of an existing
volume or decrease it.


HTH

> --
> Lindsay Mathieson
>
>


-- 
Pranith
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.gluster.org/pipermail/gluster-users/attachments/20170610/ce0584de/attachment.html>


More information about the Gluster-users mailing list