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
When a retry happens where one run is the parent of other runs, the child runs should be stopped if they were not already in a terminal state. If the retry was triggered by a failure in one of these child runs, and there was no parallelism, this will have happened naturally. However, there are two cases where a retry can be triggered where there are still active child runs:
(a) if there is parallelism in the child runs, and only one of the branches failed (other branches would still be active)
(b) if the parent had a timeout, and therefore the parent fails despite the children being otherwise healthy.
The text was updated successfully, but these errors were encountered:
augray
changed the title
'retry' should mark child runs that weren't yet terminal as failed
'retry' should mark child runs that weren't yet terminal as failed/canceled
Apr 15, 2024
This will be kind of tricky to do: currently there is an assumption that if there are any failed/canceled runs in the pipeline, the whole pipeline should be failed. Fixing this bug would mean changing that assumption.
When a retry happens where one run is the parent of other runs, the child runs should be stopped if they were not already in a terminal state. If the retry was triggered by a failure in one of these child runs, and there was no parallelism, this will have happened naturally. However, there are two cases where a retry can be triggered where there are still active child runs:
(a) if there is parallelism in the child runs, and only one of the branches failed (other branches would still be active)
(b) if the parent had a timeout, and therefore the parent fails despite the children being otherwise healthy.
The text was updated successfully, but these errors were encountered: