PipelineRun timeouts delete TaskRun related pods

[rporres]

The current implementation of the PipelineRun timeout is done through deleting the TaskRun associated pods. While this is an understandable implementation as there is no easy way to stop a pod, it makes the debugging of a PipelineRun that has failed completely impossible. It also took me a while to understand why the related pods were being deleted.

In the current shape, the PipelineRun timeout may not be very helpful for people that need to access the logs of the failed TaskRuns. It may be helpful to add a note in the docs about how timeouts are being implemented so that people choose wisely if they want to use the feature.

Since pipeline timeouts are a very desirable feature, they should be implemented in a way that keeps the underlying pods available for debugging purposes.

[vdemeester]

/kind feature

One possible exploration to do here would be to make the entrypoint handle the timeout somehow. That way, we could stop the execution of the pod/containers without deleteting it.

[afrittoli]

Related to the behaviour of "cancel" - see #3238

[tekton-robot]

Issues go stale after 90d of inactivity.
Mark the issue as fresh with /remove-lifecycle stale with a justification.
Stale issues rot after an additional 30d of inactivity and eventually close.
If this issue is safe to close now please do so with /close with a justification.
If this issue should be exempted, mark the issue as frozen with /lifecycle frozen with a justification.

/lifecycle stale

Send feedback to tektoncd/plumbing.

[rporres]

/remove-lifecycle stale

I doubt we want to close this, it is an important feature to have as it will make the timeouts really useful

[tekton-robot]

Issues go stale after 90d of inactivity.
Mark the issue as fresh with /remove-lifecycle stale with a justification.
Stale issues rot after an additional 30d of inactivity and eventually close.
If this issue is safe to close now please do so with /close with a justification.
If this issue should be exempted, mark the issue as frozen with /lifecycle frozen with a justification.

/lifecycle stale

Send feedback to tektoncd/plumbing.

[rporres]

/remove-lifecycle stale

I doubt we want to close this, it is an important feature to have as it will make the timeouts really useful

[tekton-robot]

Issues go stale after 90d of inactivity.
Mark the issue as fresh with /remove-lifecycle stale with a justification.
Stale issues rot after an additional 30d of inactivity and eventually close.
If this issue is safe to close now please do so with /close with a justification.
If this issue should be exempted, mark the issue as frozen with /lifecycle frozen with a justification.

/lifecycle stale

Send feedback to tektoncd/plumbing.

[vdemeester]

/lifecycle frozen
Should be more or less taken care of with #4618, as the timeout is cancelling the Task. Today cancel means delete, but with that proposed feature, we would not delete the pod anymore.