For sake of completeness here below I pasted the relevant part of the 
etc-glusterfs-glusterd.vol.log log file from my arbiter node. Strangely enough 
one of the log message says that the file 
/var/lib/glusterd/geo-replication/gsyncd_template.conf does not exist but I 
checked and it does exist on all nodes including the arbiter node.
[2017-07-29 14:02:02.872498] E [MSGID: 106293] 
[glusterd-geo-rep.c:676:glusterd_query_extutil_generic] 0-management: reading 
data from child failed
[2017-07-29 14:02:02.872511] E [MSGID: 106305] 
[glusterd-geo-rep.c:1944:is_geo_rep_active] 0-management: Unable to get 
configuration data for myvolume(master), ssh://gfs1geo.domain.tld::myvolume
-geo:fe090023-7add-40b2-a6c5-d7e7ea27e12a(slave)
[2017-07-29 14:02:02.872588] E [MSGID: 106300] 
[glusterd-geo-rep.c:2104:glusterd_check_geo_rep_running] 0-management: 
_get_slave_satus failed
[2017-07-29 14:02:02.872607] E [MSGID: 106301] 
[glusterd-op-sm.c:5423:glusterd_op_ac_stage_op] 0-management: Stage failed on 
operation 'Volume Stop', Status : -1
[2017-07-29 14:01:22.986242] W [MSGID: 106029] 
[glusterd-geo-rep.c:2552:glusterd_get_statefile_name] 0-management: Config file 
(/var/lib/glusterd/geo-replication/myvolume_gfs1geo.domain.tld_myvolume-geo/gsyncd.conf)
 missing. Looking for template config file 
(/var/lib/glusterd/geo-replication/gsyncd_template.conf) [No such file or 
directory]
[2017-07-29 14:01:34.270084] E [MSGID: 106459] 
[glusterd-geo-rep.c:4601:glusterd_read_status_file] 0-management: Unable to get 
status data for myvolume(master), gfs1geo.domain.tld::myvolume-geo(
slave), /srv/glusterfs/myvolume/brick(brick)

> -------- Original Message --------
> Subject: Re: [Gluster-users] Not possible to stop geo-rep after adding 
> arbiter to replica 2
> Local Time: July 29, 2017 4:07 PM
> UTC Time: July 29, 2017 2:07 PM
> From: m...@protonmail.ch
> To: Karan Sandha <ksan...@redhat.com>
> Rahul Hinduja <rhind...@redhat.com>, Kotresh Hiremath Ravishankar 
> <khire...@redhat.com>, Gluster Users <gluster-users@gluster.org>
> I managed to force stopping geo replication using the "force" parameter after 
> the "stop" but there are still other issues related to the fact that my geo 
> replication setup was created before I added the additional arbiter node to 
> my replca.
> For example when I would like to stop my volume I simply can't and I get the 
> following error:
> volume stop: myvolume: failed: Staging failed on arbiternode.domain.tld. 
> Error: geo-replication Unable to get the status of active geo-replication 
> session for the volume 'myvolume'.
> What should I do here? Should I delete the geo-replication and re-create it? 
> Will it then have to sync again all my data? I have around 500 GB of data so 
> I would like to avoid that if possible. Any other solutions?
> Thanks,
> M.
>
>> -------- Original Message --------
>> Subject: Re: [Gluster-users] Not possible to stop geo-rep after adding 
>> arbiter to replica 2
>> Local Time: July 29, 2017 12:32 PM
>> UTC Time: July 29, 2017 10:32 AM
>> From: ksan...@redhat.com
>> To: mabi <m...@protonmail.ch>, Rahul Hinduja <rhind...@redhat.com>, Kotresh 
>> Hiremath Ravishankar <khire...@redhat.com>
>> Gluster Users <gluster-users@gluster.org>
>>
>> Adding Rahul and Kothresh who are SME on geo replication
>> Thanks & Regards
>> Karan Sandha
>>
>> On Sat, Jul 29, 2017 at 3:37 PM, mabi <m...@protonmail.ch> wrote:
>>
>>> Hello
>>> To my two node replica volume I have added an arbiter node for safety 
>>> purpose. On that volume I also have geo replication running and would like 
>>> to stop it is status "Faulty" and keeps trying over and over to sync 
>>> without success. I am using GlusterFS 3.8.11.
>>> So in order to stop geo-rep I use:
>>> gluster volume geo-replication myvolume gfs1geo.domain.tld::myvolume-geo 
>>> stop
>>> but it fails to stop it as you can see in the output below:
>>> Staging failed on arbiternode.domain.tld. Error: Geo-replication session 
>>> between myvolume and gfs1geo.domain.tld::myvolume-geo does not exist.
>>> geo-replication command failed
>>> How can I now stop geo replication? Is there a manual way to do that?
>>> Regards,
>>> Mabi
>>> _______________________________________________
>>> Gluster-users mailing list
>>> Gluster-users@gluster.org
>>> http://lists.gluster.org/mailman/listinfo/gluster-users
>> --
>>
>> Regards & Thanks
>> Karan Sandha
_______________________________________________
Gluster-users mailing list
Gluster-users@gluster.org
http://lists.gluster.org/mailman/listinfo/gluster-users

Reply via email to