Bug #55224
openorchestrator deploys some containers with wrong image
0%
Description
After converting a Nautilus cluster to Pacific with cephadm, some daemons where deployed with the wrong image:
[ceph: root@alpha ~]# ceph orch ps
NAME HOST PORTS STATUS REFRESHED AGE MEM USE MEM LIM VERSION IMAGE ID CONTAINER ID
crash.alpha alpha running (26m) 19s ago 26m 33.3M - 16.2.5-387-g7282d81d 41387741ad94 4802f153b642
crash.bravo bravo running (25m) 20s ago 25m 19.4M - 16.2.5-387-g7282d81d 41387741ad94 d72788d8c6df
crash.charlie charlie running (26m) 20s ago 25m 7650k - 16.2.5-387-g7282d81d 41387741ad94 3b1b13673c8c
crash.delta delta running (25m) 20s ago 25m 30.6M - 16.2.5-387-g7282d81d 41387741ad94 b604735e0eee
crash.echo echo running (20m) 3m ago 25m 21.8M - 16.2.5-387-g7282d81d 41387741ad94 3b7a7a7bd6c0
mgr.alpha alpha running (38m) 19s ago 42m 498M - 16.2.7 c92aec2cd894 0d60b2218c48
mgr.bravo bravo running (48m) 20s ago 40m 460M - 16.2.7 c92aec2cd894 8cd13c80693b
mon.alpha alpha running (38m) 19s ago - 326M 2048M 16.2.7 c92aec2cd894 650fd0087b35
mon.bravo bravo running (50m) 20s ago - 264M 2048M 16.2.7 c92aec2cd894 d0408ee1656a
mon.charlie charlie running (51m) 20s ago - 273M 2048M 16.2.7 c92aec2cd894 00c1fbe798f3
osd.0 alpha running (38m) 19s ago 38m 68.9M 4096M 16.2.7 c92aec2cd894 83674ab68644
osd.1 bravo running (37m) 20s ago 36m 70.2M 4096M 16.2.7 c92aec2cd894 f99f1528634b
osd.2 charlie running (37m) 20s ago 36m 71.9M 4096M 16.2.7 c92aec2cd894 225816a2fd06
osd.3 delta running (35m) 20s ago 35m 41.0M 4096M 16.2.7 c92aec2cd894 cc48ad5d52b2
osd.4 echo running (4m) 3m ago 4m 41.9M 4096M 16.2.5-387-g7282d81d 41387741ad94 e1734be4678e
osd.5 echo running (4m) 3m ago 4m 36.5M 4096M 16.2.5-387-g7282d81d 41387741ad94 f06f3b9e99bb
The managers, mons and osd 0 to 3 were converted by me, and run 16.2.7 as expected. osd 4 and 5 were deployed later by ceph orch and run the wrong version as you can see. The same goes for the crash daemons.
Updated by Adam King about 2 years ago
Can you see if using the upgrade command can fix the issue. as in
ceph orch upgrade start quay.io/ceph/ceph:v16.2.7
That might get all the daemons back in line for you.
Updated by Ruben Kerkhof about 2 years ago
Adam King wrote:
Can you see if using the upgrade command can fix the issue. as in
[...]
That might get all the daemons back in line for you.
Yes, that works as a workaround.