Repository navigation
test(parallel-agents): cover the listening-port check that the two-slot test stubs out - #97
Merged
Merged
Conversation
…chine's listening ports worktree-new.sh treats a port something listens on as taken. The 2-slot test used 41080 and 41180, so any listener on either left the second worktree without a slot and failed the run. The test now stubs the listening probe in its own copy of the scripts, so only the env files reserve a port, and a new test covers the listening probe on a port it binds itself. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
…stening-port test Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
mauricios
changed the title
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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
#95 fixed the flaky
worktree-new: with 2 slots, two worktrees take both ports(a host listener on 41080 or 41180 took a slot) by stubbing the listening probe in the test's copy of_worktree-lib.sh. This PR found the same root cause independently, reproduced it by binding 41180, and was merged onto #95's stub.The stub leaves
port_in_usewith no test. Nothing checks any more thatworktree-new.shskips a port something is listening on.Change
New test:
worktree-new: skips a port something listens on. It binds a port the OS picks itself, so the port is known to be busy and owned by the test. It puts the single slot on that port and expectsno free port slot. It's deterministic: no fixed port, no dependence on the host.Only
evals/changes, so there's no CHANGELOG entry or plugin rebuild.Verification
evals/static/test-modules.shran 30 times in a row with 41080 and 41180 both held by listeners, with 0 failures../evals/run-evals.sh, all static suites passed.🤖 Generated with Claude Code