Fix hang on shutdown in test_refork [changelog skip] #2442
Merged
Conversation
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
If a worker receives a `TERM` signal before the server has started, prevent it from starting at all by clearing `restart_server`.
When `fork_worker` is used, the call to `Process.wait` in `wait_workers` returns `ECHILD` for forked workers. Add a call to `w.term` in the rescue clause to match the behavior when `fork_worker` is not used, ensuring workers shut down properly.
Unit test clarifying behavior of commands sent before Server#run is called.
8 tasks
nateberkopec
added
bug
maintenance
waiting-for-review
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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Outdated suggestions cannot be applied.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.
Description
This PR fixes a hang on shutdown that occurred intermittently in
test_refork(example). I reproduced the hang more consistently runningtest_reforkwith an extrasleepafter the'b'(boot) message and before it starts the server (e.g., line 106 below):puma/lib/puma/cluster/worker.rb
Lines 99 to 108 in aae4d4d
The PR contains two small fixes.
TERMsignal (0018edd). This handles some (but not 100% of all) edge cases where theTERMsignal is received and callsserver.stopbefore the server has started.fork_workeris used, the call toProcess.waitinwait_workersreturnsECHILDfor forked workers. This commit adds a call tow.termin the rescue clause, matching the behavior whenfork_workeris not used and ensuring workers shut down properly. This should handle all other edge-cases where a worker receivesTERMbefore the server starts.Finally, note that this bug started occurring more frequently after #2435, because that PR changed behavior so that calling
#stopbefore#runnow has no effect. I've addedtest_command_ignored_before_runto clarify this new behavior in a unit test. I think this is more intuitive and predictable behavior overall, but we can change it back if anyone disagrees or if the change continues to cause any other issues.