retrofitting failure is not reflected to juju unit status

[Nobuto Murata]

When another bug 2037935 happened, I realized that juju unit status is still in maintenance/idle although the retrofitting process clearly failed.

  octavia-diskimage-retrofit/0* maintenance idle 192.168.151.105 Retrofitting auto-sync/ubuntu-jammy-22.04-amd64-server-20230927-disk1.img

It would be nice the unit status turns into error/blocked when the retrofitting process failed instead of keeping maintenance/Retrofitting status.

$ juju show-operation 13
{}
ERROR operation 13 not found
root@quick-maas:~# juju show-operation -m openstack 13
summary: retrofit-image run on octavia-diskimage-retrofit/leader
status: failed
action:
  name: retrofit-image
  parameters:
    force: false
    source-image: ""
timing:
  enqueued: 2023-10-04 05:36:06 +0000 UTC
  started: 2023-10-04 05:36:06 +0000 UTC
  completed: 2023-10-04 05:38:11 +0000 UTC
tasks:
  "14":
    host: octavia-diskimage-retrofit/0
    status: failed
    timing:
      enqueued: 2023-10-04 05:36:06 +0000 UTC
      started: 2023-10-04 05:36:06 +0000 UTC
      completed: 2023-10-04 05:38:11 +0000 UTC
    message: Command '['octavia-diskimage-retrofit', '/var/snap/octavia-diskimage-retrofit/common/tmp/tmpomjo03dt',
      '/var/snap/octavia-diskimage-retrofit/common/tmp/tmphzjhxmvi']' returned non-zero
      exit status 1.
    results:
      return-code: 0