On Tue, Feb 12, 2013 at 4:48 AM, Holcombe, Christopher <cholcomb@xxxxxxxxxxx> wrote: > I started a reweight operation yesterday. Before it finished I changed the crush map to to reweight all the osd's at once. The cluster seemed to come back ok but 27 pgs are stuck unclean and I'm not sure what to do at this point. I knew I should have waited for that reweight to complete... > I couldn't find much information on the wiki other than stuck unclean can happen. Any tips for how I can recover from this? I'm running Ubuntu 12.04 with 0.56.1 That's a bit odd. Can you check the mapping for the stuck PGs and see if they have the correct counts?[1] And check "ceph osd tree" to make sure that you actually updated the weights on all your OSDs? -Greg [1]: http://ceph.com/docs/master/rados/operations/monitoring-osd-pg/#identifying-troubled-pgs Dump the stuck ones, then "ceph pg dump" and grep for the ones which are stuck. _______________________________________________ ceph-users mailing list ceph-users@xxxxxxxxxxxxxx http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
 |