Do something like this after reconnecting the brick without attrs and
.glusterfs to those
cd BRICKPATH
sudo find . -path ./.glusterfs -prune -o -exec stat 'MUNTPATH/{}' \;
if runs stat for each filename in the brick through mountpoint,
excluding .glusterfs folder
On Sun, Feb 25, 2018 at 3:16 PM, Mitja Mihelič <mitja.mihelic@xxxxxxxx> wrote:
> Let me see if I understand this.
> Remove attrs from the brick and delete the .glusterfs folder. Data stays in
> place. Add the brick to the volume.
> Since most of the data is the same as on the actual volume it does not need
> to be synced, and the heal operation finishes much faster.
> Do I have this right?
>
> Kind regards,
> Mitja
>
>
> On 25/02/2018 17:02, Vlad Kopylov wrote:
>>
>> .gluster and attr already in that folder so it would not connect it as a
>> brick
>> I don't think there is option to "reconnect brick back"
>>
>> what I did many times - delete .gluster and reset attr on the folder,
>> connect the brick and then update those attr. with stat
>> commands example here
>> http://lists.gluster.org/pipermail/gluster-users/2018-January/033352.html
>>
>> Vlad
>>
>> On Sun, Feb 25, 2018 at 4:37 AM, Mitja Mihelič <mitja.mihelic@xxxxxxxx>
>> wrote:
>>>
>>> Hi!
>>>
>>> I am running a replica 3 volume. On server2 I wanted to move the brick to
>>> a
>>> new disk.
>>> I removed the brick from the volume:
>>> gluster volume remove-brick VOLUME rep 2
>>> server2:/gluster/VOLUME/brick0/brick force
>>>
>>> I unmounted the old brick and mounted the new disk to the same location.
>>> I added the empty new brick to the volume:
>>> gluster volume add-brick VOLUME rep 3
>>> server2:/gluster/VOLUME/brick0/brick
>>>
>>> There is about 2TB of data on the volume and they are all small files,
>>> photos and documents. The whole heal process is rather slow. In 10 hours
>>> only 120GB was copied to the new brick.
>>> I remove the new brick:
>>> gluster volume remove-brick VOLUME rep 2
>>> server2:/gluster/VOLUME/brick0/brick force
>>>
>>> I remounted the old brick in its place. Now I would like to re-add the
>>> existing old brick back to the volume, so I have a replica 3 again.
>>> But the add operation fails with:
>>> volume add-brick: failed: /gluster/VOLUME/brick0/brick is already part of
>>> a
>>> volume
>>>
>>> How can I re-add this brick back to the volume, and how save is it? Can I
>>> use force to add it?
>>> There has been little to no data change on the volume since the old brick
>>> was removed in the first place.
>>>
>>> Kind regards,
>>> Mitja
>>> _______________________________________________
>>> Gluster-users mailing list
>>> Gluster-users@xxxxxxxxxxx
>>> http://lists.gluster.org/mailman/listinfo/gluster-users
>
>
_______________________________________________
Gluster-users mailing list
Gluster-users@xxxxxxxxxxx
http://lists.gluster.org/mailman/listinfo/gluster-users
