[Gluster-infra] slave21 and 23 offline due to disk full

[Michael Scherer]

Le mercredi 09 septembre 2015 à 16:51 +0200, Michael Scherer a écrit :
> Hi,
> 
> just found out that slave21 and 23 were  offline due to their disk being
> full. The issue is 14G of log, due to something creating tarball
> of /var/log/glusterfs, and place the tarball in /var/log/glusterfs/ 

So slave 23 was down tonight, something something about ext4 corruption
in log and the salt-minion unable to start, complaining about /var/cache
I/O error. 
I did reboot, didn't help, cleaned /var/cache/salt, minion is back. 

Not sure if anything else got corrupted, but I am a bit concerned that
this could happen in the first place.
-- 
Michael Scherer
Sysadmin, Community Infrastructure and Platform, OSAS

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 836 bytes
Desc: This is a digitally signed message part
URL: <http://www.gluster.org/pipermail/gluster-infra/attachments/20150909/41309993/attachment.sig>