[feature request] Support progressDeadlineSeconds in CloneSet

[ronaknnathani]

What would you like to be added:
Add support for progressDeadlineSeconds in CloneSet. .spec.progressDeadlineSeconds is an optional field in Deployment that specifies the number of seconds one wants to wait for their Deployment to progress before the system reports back that the Deployment has failed progressing - surfaced as a condition with type: Progressing, status: "False" and reason: ProgressDeadlineExceeded in the status of the resource.

Reference: https://kubernetes.io/docs/concepts/workloads/controllers/deployment/#progress-deadline-seconds

Why is this needed:
In the case of CloneSets, if the rollout is stuck for any reason - image pull errors, insufficient permissions, insufficient capacity, unhealthy app, etc. - currently there is no way to identify that the rollout is stuck by looking at the CloneSet status. With a feature like progressDeadlineSeconds, one can specify a time and if the rollout hasn't made any progress for that period, the CloneSet status can be updated to reflect that. The CloneSet controller would still continue to reconcile the requests, however, this status allows continuous delivery pipelines and any other controllers that build on top of CloneSet to know that the rollout failed.

[ronaknnathani]

@furykerry do we need a proposal for this or do you think we can send a PR for review?

[furykerry]

yes, plz draft a proposal for this feature in https://github.com/openkruise/kruise/blob/master/docs/proposals

[furykerry]

@ronaknnathani we're drafting a proposal for this feature, however the definition of progressDeadlineSeconds can be different from Deployment because cloneset support partition. Can you give us some feedback about the potential use case of progressDeadlineSeconds when partition is set ?

[stale]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.