Problem is now solved, the cluster is now backfilling/recovering normally, no more NEAR FULL OSD.
It turns out that I have RBD objects that should have been deleted long time ago but it's still there. Openstack Glance did not removed it, I think it's an issue with snapshots, an RBD file can't be deleted unless it's snapshots are purged. So I compared all my glance images to the RBD counterpart and identified which are not there and deleted them. So from 81% utilization I am down to 61%. --- [root@controller-node opt]# ceph df GLOBAL: SIZE AVAIL RAW USED %RAW USED 100553G 39118G 61435G 61.10 POOLS: NAME ID USED %USED OBJECTS images 4 1764G 1.76 225978 volumes 5 18533G 18.43 4762609 [root@controller-node opt]# --- On Sat, Feb 20, 2016 at 5:38 AM, Lionel Bouton < lionel-subscript...@bouton.name> wrote: > Le 19/02/2016 17:1 > > 7, Don Laursen a écrit : > > Thanks. To summarize > > Your data, images+volumes = 27.15% space used > > Raw used = 81.71% used > > > > This is a big difference that I can’t account for? Can anyone? So is your > cluster actually full? > > > I believe this is the pool size being accounted for and it is harmless: 3 > x 27.15 = 81.45 which is awfully close to 81.71. > We have the same behavior on our Ceph cluster. > > > > I had the same problem with my small cluster. Raw used was about 85% and > actual data, with replication, was about 30%. My OSDs were also BRTFS. > BRTFS was causing its own problems. I fixed my problem by removing each OSD > one at a time and re-adding as the default XFS filesystem. Doing so brought > the percentages used to be about the same and it’s good now. > > > That's odd : AFAIK we had the same behaviour with XFS before migrating to > BTRFS. > > Best regards, > > Lionel > > _______________________________________________ > ceph-users mailing list > ceph-users@lists.ceph.com > http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com > > ᐧ
_______________________________________________ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com