The reason why you moved to ceph storage, is that you do not want to do such things. Remove the drive, and let ceph recover.
On May 6, 2019 11:06 PM, Florent B wrote: > > Hi, > > It seems that OSD disk is dead (hardware problem), badblocks command > returns a lot of badblocks. > > Is there any way to force Ceph to recover as many objects as it can on > that disk ? I know data will be lost but I would like to recover as many > as possible. > > Thank you. > > On 04/05/2019 18:56, Florent B wrote: > > Hi, > > > > I have a problem with an OSD that stopped itself and then can't restart. > > > > I use Luminous 12.2.12. > > > > I can see these lines in logs : > > > > -2> 2019-05-04 18:48:33.687087 7f3aedfe1700 4 rocksdb: EVENT_LOG_v1 > > {"time_micros": 1556988513687079, "job": 3, "event": > > "compaction_started", "files_L1": [7456], "files_L2": [7414, 7415, 7416, > > 7417, 7418, 7419, 7420], "score": 1.25456, "input_data_size": 494084689} > > -1> 2019-05-04 18:48:33.947802 7f3b023bde00 -1 bdev(0x55ed83fb3200 > > /var/lib/ceph/osd/ceph-38/block) direct_read_unaligned > > 0x3cbb14c2b8~1c9631 error: (61) No data available > > 0> 2019-05-04 18:48:33.952965 7f3b023bde00 -1 > > /build/ceph-12.2.12/src/os/bluestore/BlueFS.cc: In function 'int > > BlueFS::_read_random(BlueFS::FileReader*, uint64_t, size_t, char*)' > > thread 7f3b023bde00 time 2019-05-04 18:48:33.947917 > > /build/ceph-12.2.12/src/os/bluestore/BlueFS.cc: 936: FAILED assert(r == 0) > > > > Is there any way to recover this problem ? > > > > Thank you. > > > > Florent > > > > _______________________________________________ > > 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
_______________________________________________
ceph-users mailing list
ceph-users@xxxxxxxxxxxxxx
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
