Is there a 0.138_head directory under current/ on any of your osds? If so, can you post the log to that osd? I could also use the osd.0 log. -Sam On Wed, Apr 4, 2012 at 2:44 PM, Damien Churchill <damoxc@xxxxxxxxx> wrote: > I've uploaded them to: > > http://damoxc.net/ceph/osdmap > http://damoxc.net/ceph/pg_dump > > Thanks > > On 4 April 2012 21:51, Samuel Just <sam.just@xxxxxxxxxxxxx> wrote: >> Can you post a copy of your osd map and the output of 'ceph pg dump' ? >> You can get the osdmap via 'ceph osd getmap -o <filename>'. >> -Sam >> >> On Wed, Apr 4, 2012 at 1:12 AM, Damien Churchill <damoxc@xxxxxxxxx> wrote: >>> Hi, >>> >>> I'm having some trouble getting some pgs to stop being inactive. The >>> cluster is running 0.44.1 and the kernel version is 3.2.x. >>> >>> ceph -s reports: >>> 2012-04-04 09:08:57.816029 pg v188540: 990 pgs: 223 inactive, 767 >>> active+clean; 205 GB data, 1013 GB used, 8204 GB / 9315 GB avail >>> 2012-04-04 09:08:57.817970 mds e2198: 1/1/1 up {0=node24=up:active}, >>> 4 up:standby >>> 2012-04-04 09:08:57.818024 osd e5910: 5 osds: 5 up, 5 in >>> 2012-04-04 09:08:57.818201 log 2012-04-04 09:04:03.838358 osd.3 >>> 172.22.10.24:6801/30000 159 : [INF] 0.13d scrub ok >>> 2012-04-04 09:08:57.818280 mon e7: 3 mons at >>> {node21=172.22.10.21:6789/0,node22=172.22.10.22:6789/0,node23=172.22.10.23:6789/0} >>> >>> ceph health says: >>> 2012-04-04 09:09:01.651053 mon <- [health] >>> 2012-04-04 09:09:01.666585 mon.1 -> 'HEALTH_WARN 223 pgs stuck >>> inactive; 223 pgs stuck unclean' (0) >>> >>> I was wondering if anyone has any suggestions about how to resolve >>> this, or things to look for. I've tried restarted the ceph daemons on >>> the various nodes a few times to no-avail. I don't think that there is >>> anything wrong with any of the nodes either. >>> >>> Thanks in advance, >>> Damien >>> -- >>> 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 -- 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
