Skip to content

Conversation

@jder
Copy link
Contributor

@jder jder commented Apr 5, 2025

There seems to be a bug where cancelling a workflow doesn't take down the runner. In general 3 days is a long time to leave something idle, so let's be a bit more aggressive about shutting them down while we work on fixing the bug.

(Edit: sorry, wrong org)

There seems to be a bug where cancelling a workflow doesn't take down
the runner. In general 3 days is a long time to leave something idle,
so let's be a bit more aggressive about shutting them down while we
work on fixing the bug
@jder jder closed this Apr 5, 2025
@jder jder deleted the u/jder/better-fallback branch April 5, 2025 14:42
@jder jder restored the u/jder/better-fallback branch April 5, 2025 14:44
@ravwojdyla
Copy link
Collaborator

ravwojdyla commented Apr 5, 2025

@jder np. Btw, cancelling a workflow should cleanup the worker via #34 (hopefully that's in your org).

@jder
Copy link
Contributor Author

jder commented Apr 5, 2025

Thanks for letting me know! Somehow we ended up with 10 idle workers and no actions running and the main thing that's changed recently is that we started cancelling stale runs after new pushes. I'll try to repro and investigate on monday.

@jder
Copy link
Contributor Author

jder commented Apr 7, 2025

FYI after digging into it, the two root causes for VMs hanging around were #10 and the thing resolved in #60.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants