On Tue, Jan 15, 2019 at 3:51 PM Sergei Shvarts <storm@xxxxxxxxxxxx> wrote: > > Hello ceph users! > > A couple of days ago I've got a ceph health error - mds0: Metadata damage detected. > Overall ceph cluster is fine: all pgs are clean, all osds are up and in, no big problems. > Looks like there is not much information regarding this class of issues, so I'm writing this message and hope somebody can help me. > > here is the damage itself > ceph tell mds.0 damage ls > 2019-01-15 07:47:04.651317 7f48c9813700 0 client.312845186 ms_handle_reset on 192.168.0.5:6801/1186631878 > 2019-01-15 07:47:04.656991 7f48ca014700 0 client.312845189 ms_handle_reset on 192.168.0.5:6801/1186631878 > [{"damage_type":"dir_frag","id":3472877204,"ino":1100954978087,"frag":"*","path":"\/public\/video\/3h\/3hG6X7\/screen-msmall"}] > Looks like object 1005607c727.00000000 in cephfs metadata pool is corrupted. please run following commands and send mds.0 log to us ceph tell mds.0 injectargs '--debug_mds 10' ceph tell mds.0 damage rm 3472877204 ls <cephfs mount point>/public/video/3h/3hG6X7/screen-msmall ceph tell mds.0 injectargs '--debug_mds 0' Regards Yan, Zheng > Best regards, > Sergei Shvarts > _______________________________________________ > 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