On Fri, Jan 15, 2021 at 4:36 AM Rafael Diaz Maurin
<Rafael.DiazMaurin(a)univ-rennes1.fr> wrote:
Hello cephers,
I run Nautilus (14.2.15)
Here is my context : each night a script take a snapshot from each RBD volume in a pool
(all the disks of the VMs hosted) on my ceph production cluster. Then each snapshot is
exported (rbd export-diff | rbd import-diff in SSH) towards my ceph bakup cluster.
0. Yesterday, and for the firt time, I had this error when I ran ''rbd -p MY_POOL
du'' on the ceph backup cluster :
021-01-14 11:41:29.294 7f1f2fbcf0c0 -1 librbd::DiffIterate: diff_object_map: failed to
load object map rbd_object_map.781776950f8dd5.000000000000b740
1. I tried to find the image concerned by the error, and I find this one, with a slight
diffence with the number of the object-map :
rbd image 'MY_IMAGE':
size 30 GiB in 7680 objects
order 22 (4 MiB objects)
snapshot_count: 32
id: 781776950f8dd5
block_name_prefix: rbd_data.781776950f8dd5
format: 2
features: layering, exclusive-lock, object-map, fast-diff, deep-flatten, operations
op_features: snap-trash
flags:
create_timestamp: Thu Mar 19 01:00:27 2020
access_timestamp: Thu Mar 19 01:00:27 2020
modify_timestamp: Thu Jan 14 01:42:23 2021
2. Then I ran a check wich not show any error
rbd -p MY_POOL object-map check "MY_IMAGE"
Object Map Check: 100% complete...done.
3. Then, I still ran a rebuild
rbd -p MY_POOL object-map rebuild "MY_IMAGE"
Object Map Rebuild: 100% complete...done.
4. But the error is still here :
2021-01-15 09:33:58.775 7fa088e350c0 -1 librbd::DiffIterate: diff_object_map: failed to
load object map rbd_object_map.781776950f8dd5.000000000000b740
You would need to rebuild the snapshot object map. First find the name
of the snapshot associated with snapshot id b740 (46912) via "rbd snap
ls <image-spec>", then run rebuild on that particule snapshot via "rbd
object-map rebuild <image-spec>@<snap name>".
Do someone know how to find the the image concerned by the object-map error and fix this
error.
Thank you for your help,
Rafael
_______________________________________________
ceph-users mailing list -- ceph-users(a)ceph.io
To unsubscribe send an email to ceph-users-leave(a)ceph.io
--
Jason