[Bugs] [Bug 858732] glusterd does not start anymore on one node

[bugzilla at redhat.com]

https://bugzilla.redhat.com/show_bug.cgi?id=858732

Niels de Vos <ndevos at redhat.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |Triaged
                 CC|                            |kparthas at redhat.com,
                   |                            |ndevos at redhat.com
              Flags|needinfo?(bugs at gluster.org) |needinfo?(kparthas at redhat.c
                   |                            |om)



--- Comment #9 from Niels de Vos <ndevos at redhat.com> ---
(In reply to Anatoly Belikov from comment #8)
> Any update on this bug? We have run into the same problem:
...
> [2015-05-18 10:38:45.142777] E [run.c:190:runner_log] 0-glusterd: command
> failed: /usr/lib/x86_64-linux-gnu/glusterfs/gsyncd -c
> /etc/glusterd/geo-replication/gsyncd.conf --config-set-rx gluster-params
> xlator-option=*-dht.assert-no-child-down=true .
...

This is not the same problem, but it has the same effect. Please file a new bug
report against the geo-replication component. This could be a packaging issue,
or something else that caused the "gsync" command to fail.


---

The original problem reported in this bug is caused by glusterd being unable to
read some of its configuration files. This can (or could?) happen when /var/lib
is full or out of inodes. Cleanup and manually restoring the configuration
under /var/ilb/glusterd is needed in that case. KP or some of the other
GlusterD developers can chime in with more details, and maybe a link to the
documentation or email that describes how to restore the configuration.

-- 
You are receiving this mail because:
You are on the CC list for the bug.
You are the assignee for the bug.