Hi, On Thu, 28 May 2015, Ugis wrote: > Hi! > > I have been watching changes in "ceph -s" output for a while and > noticed that in this line: > 3324/7888981 objects degraded (0.042%); 1995972/7888981 objects > misplaced (25.301%) > rather misplaced object count drops constantly, but degraded object > count drops just occasionally. > > Quick googling did not reveal any mention of recovery/rebalancing > policy(except client vs recovery io). > Is there any? > I mean - it would be reasonable to recover degraded objects in first > place and only then to rebalance misplaced objects as that would > reduce risk of losing data. > > That could be default recovey policy, but if other policy needed - > that could be specifyable via some recovery variable, like it is now > in recovery section here: > http://dachary.org/loic/ceph-doc/rados/configuration/osd-config-ref/ > > P.S. no active client I/O currently present. If the recovery and rebalancing ever contend for the same node, recovery is prioritized. What is probaly happening in your case is that the rebalancing is happening elsewhere in the cluster where nothing is degraded... sage -- To unsubscribe from this list: send the line "unsubscribe ceph-devel" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
