Christian, Still running Dumpling (I know I have to start upgrading). Cluster has 66 OSD’s and a total size close to 100 GB. Cluster is running for around 2 years now and the monitor server has an uptime of 258 days. The LOG file is 1.2 GB in size and ls shows the current time for it. The LOG.OLD is 1.5 GB and ls tells me a date of Jan 23 2015 (which is 258 before today). lsof tells that the LOG is open, while the LOG.OLD isn’t. So it don’t seems to rotate, without restarting the monitor. Regards, Erwin > Op 8 okt. 2015, om 09:57 heeft Christian Balzer <chibi@xxxxxxx> het volgende geschreven: > > > Hello, > > On Thu, 8 Oct 2015 09:38:02 +0200 Erwin Lubbers wrote: > >> Hi, >> >> In the /var/lib/ceph/mon/ceph-l16-s01/store.db/ directory there are two >> very large files LOG and LOG.OLD (multiple GB's) and my diskspace is >> running low. Can I safely delete those files? >> > That sounds odd, what version of Ceph are you running, something > pre-Firefly? > > How long has this monitor been running, what size, how busy is your > cluster? > > Those leveldb log files shouldn't get that big and getting auto-rotated. > > In general, if lsof shows that file being used (most likely if the date of > LOG is current) it obviously isn't safe to remove it. > > Christian > >> Regards, >> Erwin >> _______________________________________________ >> ceph-users mailing list >> ceph-users@xxxxxxxxxxxxxx >> http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com >> > > > -- > Christian Balzer Network/Systems Engineer > chibi@xxxxxxx Global OnLine Japan/Fusion Communications > http://www.gol.com/ _______________________________________________ ceph-users mailing list ceph-users@xxxxxxxxxxxxxx http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
 |