Repository navigation
Retry an auto-fix once when the first run is lost - #370
Merged
Merged
Conversation
The first auto-fix run for the Keycloak PR died when GitHub's runner lost contact, and the once-per-head rule then left the PR stuck. Auto-fix now allows two attempts per head and set of findings, which covers a lost run while still stopping a fix the job cannot make from looping. 💘 Generated with Crush Assisted-by: Crush:claude-opus-5-5 Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Why
Auto-fix sent the implementation job to the Keycloak PR (#359). That run died about an hour in with "The hosted runner lost communication with the server", before it pushed anything. Auto-fix only asked once per head commit and set of findings, so it would never have asked again, and the PR would have stayed stuck until a person noticed. I restarted that run by hand.
Fix
Auto-fix now allows two attempts per head and set of findings. The second attempt covers a run that was lost with its runner. The limit still stops a fix the job can't make from running in a loop. The PR comment says which attempt it is. A new push or new findings start the count over, as before.
Test plan
make cion9ec848bscripts/automerge_test.sh: one earlier attempt means a retry labelled "attempt 2 of 2"; two attempts mean it moves on to the next PR; changed findings ask again💘 Generated with Crush