Ceph » Orchestrator

It would be helpful for the upgrade process to expose the timings relating to the upgrade
- start time
- pause time
- active time (start time - pause time)

In addition, the current output shows services_complete (list) - but you don't know the order of the upgrade. Perhaps the instead of a list this could be a dict of the services with a pending, active, complete state to provide a sense of progress instead of watching the cephadm log? For extra visibility, of the state (eg. active) include the daemon count the admin would be more aware of where the upgrade was.
e.g.
upgrade_detail = {
mgr: complete 2/2
mon: complete 5/5
osd: active 3/1968
mds: pending
rgw: pending
prometheus: pending
<etc>
}