Actions
Bug #45418
closedcephadm: `orch reconfig` does not reconfig the container image
% Done:
0%
Source:
Tags:
Backport:
Regression:
No
Severity:
3 - minor
Reviewed:
Description
define a custom container image:
ceph config set client.nfs container_image registry.opensuse.org/filesystems/ceph/master/upstream/images/ceph/ceph:latest
`orch reconfig` of the service still uses the old image:
$ ceph config set client.nfs container_image registry.opensuse.org/filesystems/ceph/master/upstream/images/ceph/ceph:latest $ ceph orch reconfig nfs.foo Reconfigured nfs.foo.host1 on host 'host1' $ ceph orch ls --refresh NAME RUNNING REFRESHED AGE PLACEMENT IMAGE NAME IMAGE ID crash 1/1 2s ago 119m * docker.io/ceph/daemon-base:latest-master-devel 7bb50d3c5c9a mds.a 1/2 2s ago 114m count:2 docker.io/ceph/daemon-base:latest-master-devel 7bb50d3c5c9a nfs.foo 1/1 2s ago 69m count:1 docker.io/ceph/daemon-base:latest-master-devel faea5c7dc277
Updated by Michael Fritch about 4 years ago
applies to any changes to the systemd unit, unit.run, unit.poststop scripts etc.
workaround is to completely remove and deploy the service again.
relates to:
https://tracker.ceph.com/issues/45417
where the unit.poststop script needed to be updated
Updated by Michael Fritch about 4 years ago
- Status changed from New to Fix Under Review
Updated by Michael Fritch about 4 years ago
- Status changed from Fix Under Review to Rejected
confirmed that the needed functionality is already provided via the 'redeploy' command
$ ceph orch redeploy nfs.foo
Actions