Hi all, I seem to have forgotten to mention my setup. I have Ceph Hammer (ceph version 0.94.7 (d56bdf93ced6b80b07397d57e3fa68fe68304432) CentOS 7.2 w/ Linux 4.4.13 The pool in question is an EC pool on SSD with an SSD cache pool in front for RBD. I've done some more digging and I really don't understand why my pg isn't starting. The EC pool is configured for k+m of 7+3, and as far as I can tell from pg query I have 7 parts working. pgid osd last_update 64.es6 27 137824'49444 64.es5 91 137824'49444 64.es7 112 137824'49444 64.es3 189 137824'49444 64.es9 207 137824'49444 64.es8 210 137824'49444 64.es2 231 137824'49444 I have no idea why ceph won't just rebuild the missing 3 parts of the ec pool. They are listed as existing but having a last_update of 0 and no objects etc. Does anyone have any ideas? I've attached the latest version of the pg query for this pg. Thanks a lot. - HP On Sat, 2016-07-16 at 19:56 +0200, Hein-Pieter van Braam wrote: > Hi all, > > I had a crash of some OSDs today, every primary OSD of a particular > PG > just started to crash. I have recorded the information for a > bugreport. > I had reweighted the affected osds to 0 and put the processes in a > restart loop and eventually all but one placement group ended up > recovering. I have attached the output of the pg query for this pg. > > Help would be very much appreciate, I have no idea what else to do. > > Thanks, > > - HP > _______________________________________________ > ceph-users mailing list > ceph-users@xxxxxxxxxxxxxx > http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Attachment:
64.e.gz
Description: application/gzip
_______________________________________________ ceph-users mailing list ceph-users@xxxxxxxxxxxxxx http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
