You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
What would you like to be added:
We can configure a backoff limit for pods evicted due to a PodsReadyTimeout, but would like the ability to also configure a backoff limit for other types of eviction.
Why is this needed:
We have users submitting ML training jobs to Kueue, which they then track. The default preemption behavior (retry indefinitely) will be somewhat confusing for these users to track/understand, so we would like the ability to deactivate jobs on preemption.
Completion requirements:
Design doc
API change
Docs update
The artifacts should be linked in subsequent comments.
The text was updated successfully, but these errors were encountered:
What would you like to be added:
We can configure a backoff limit for pods evicted due to a
PodsReadyTimeout
, but would like the ability to also configure a backoff limit for other types of eviction.Why is this needed:
We have users submitting ML training jobs to Kueue, which they then track. The default preemption behavior (retry indefinitely) will be somewhat confusing for these users to track/understand, so we would like the ability to deactivate jobs on preemption.
Completion requirements:
The artifacts should be linked in subsequent comments.
The text was updated successfully, but these errors were encountered: