Found that this specific gfid was not pointing to any file.
Checked this with gfid resolver script
https://gist.github.com/semiosis/4392640

Moved the gfid out of gluster and all ok now.

Thanx,
Alex

On Sun, Sep 17, 2017 at 11:31 AM, Alex K <[email protected]> wrote:

> I am using gluster 3.8.12, the default on Centos 7.3
> (I will update to 3.10 at some moment)
>
> On Sun, Sep 17, 2017 at 11:30 AM, Alex K <[email protected]> wrote:
>
>> Hi all,
>>
>> I have a replica 3 with 1 arbiter.
>>
>> I see the last days that one file at a volume is always showing as
>> needing healing:
>>
>> gluster volume heal vms info
>> Brick gluster0:/gluster/vms/brick
>> Status: Connected
>> Number of entries: 0
>>
>> Brick gluster1:/gluster/vms/brick
>> Status: Connected
>> Number of entries: 0
>>
>> Brick gluster2:/gluster/vms/brick
>> *<gfid:66d3468e-00cf-44dc-a835-7624da0c5370>*
>> Status: Connected
>> Number of entries: 1
>>
>> While no split brain is reported:
>>
>> [root@v0 ~]# gluster volume heal vms info split-brain
>> Brick gluster0:/gluster/vms/brick
>> Status: Connected
>> Number of entries in split-brain: 0
>>
>> Brick gluster1:/gluster/vms/brick
>> Status: Connected
>> Number of entries in split-brain: 0
>>
>> Brick gluster2:/gluster/vms/brick
>> Status: Connected
>> Number of entries in split-brain: 0
>>
>>
>> Even when initiating the heal on the volume the file remains needing
>> heal.
>> Any idea why has this stuck?
>>
>> Thanx,
>> Alex
>>
>>
>
_______________________________________________
Gluster-users mailing list
[email protected]
http://lists.gluster.org/mailman/listinfo/gluster-users

Reply via email to