Is your feature request related to a problem?
#4131 (comment) is just the latest of several reports, where Rancher minor upgrades have downstream fleet-agent stuck in an old version.
While the upgrade checklist explicitly states that adding a grace period is recommended, we are lacking to give users (visual) feedback about version inconsistencies between upstream and downstream.
Solution you'd like
Ultimately we need to find the root cause why upgrades sometime get stuck downstream.
Until then, we should improve the feedback we give to users about the state of an ongoing upgrade, like
- a cli command to see "where's my cluster on its upgrade journey"
- a UI status / event to visualize upgrade steps
- a UI status message to indicate when downstream clusters (resp. their fleet-agent) is stuck at an old version
- ...
Alternatives you've considered
No response
Anything else?
No response
Is your feature request related to a problem?
#4131 (comment) is just the latest of several reports, where Rancher minor upgrades have downstream fleet-agent stuck in an old version.
While the upgrade checklist explicitly states that adding a grace period is recommended, we are lacking to give users (visual) feedback about version inconsistencies between upstream and downstream.
Solution you'd like
Ultimately we need to find the root cause why upgrades sometime get stuck downstream.
Until then, we should improve the feedback we give to users about the state of an ongoing upgrade, like
Alternatives you've considered
No response
Anything else?
No response