Hi, I use ceph 0.94 from wheezy repro (deb http://eu.ceph.com/debian-hammer wheezy main) inside jessie. 0.94.1 are installable without trouble, but an upgrade to 0.94.2 don't work correctly: dpkg -l | grep ceph ii ceph 0.94.1-1~bpo70+1 amd64 distributed storage and file system ii ceph-common 0.94.2-1~bpo70+1 amd64 common utilities to mount and interact with a ceph storage cluster ii ceph-fs-common 0.94.2-1~bpo70+1 amd64 common utilities to mount and interact with a ceph file system ii ceph-fuse 0.94.2-1~bpo70+1 amd64 FUSE-based client for the Ceph distributed file system ii ceph-mds 0.94.2-1~bpo70+1 amd64 metadata server for the ceph distributed file system ii libcephfs1 0.94.2-1~bpo70+1 amd64 Ceph distributed file system client library ii python-cephfs 0.94.2-1~bpo70+1 amd64 Python libraries for the Ceph libcephfs library This is the reason, why I switched back to wheezy (and clean 0.94.2) but than all OSDs on that node failed to start. Switching back to the jessie-system-disk don't solve this ploblem, because only 3 OSDs started again... My conclusion is, if now die one of my (partly brocken) jessie osd-node (like failed system ssd) I need less than an hour for a new system (wheezy), around two ours to reinitilize all OSDs (format new, install ceph) and around two days to refill the whole node. Udo Am 23.07.2015 13:21, schrieb Haomai Wang: > Do you use upstream ceph version previously? Or do you shutdown > running ceph-osd when upgrading osd? > > How many osds meet this problems? > > This assert failure means that osd detects a upgraded pg meta object > but failed to read(or lack of 1 key) meta keys from object. > > On Thu, Jul 23, 2015 at 7:03 PM, Udo Lembke <ulembke@xxxxxxxxxxxx> wrote: >> Am 21.07.2015 12:06, schrieb Udo Lembke: >>> Hi all, >>> ... >>> >>> Normaly I would say, if one OSD-Node die, I simply reinstall the OS and ceph and I'm back again... but this looks bad >>> for me. >>> Unfortunality the system also don't start 9 OSDs as I switched back to the old system-disk... (only three of the big >>> OSDs are running well) >>> >>> What is the best solution for that? Empty one node (crush weight 0), fresh reinstall OS/ceph, reinitialise all OSDs? >>> This will take a long long time, because we use 173TB in this cluster... >>> >>> >> >> Hi, >> answer myself if anybody has similiar issues and find the posting. >> >> Empty the whole nodes takes too long. >> I used the puppet wheezy system and have to recreate all OSDs (in this case I need to empty the first blocks of the >> journal before create the OSD again). >> >> >> Udo >> _______________________________________________ >> ceph-users mailing list >> ceph-users@xxxxxxxxxxxxxx >> http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com > > > _______________________________________________ ceph-users mailing list ceph-users@xxxxxxxxxxxxxx http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
 |