shahar1 opened a new pull request, #73759: URL: https://github.com/apache/airflow/pull/73759
related: #73397 This PR fixes the flaky test test_trigger_log: the loop now breaks as soon as both status lines arrive, and the upper bound is raised to 300 attempts (30 seconds) so a slow CI runner has enough headroom. <details><summary>AI Summary</summary> `test_trigger_log` gave the forked trigger runner a fixed budget of thirty `_service_subprocess(0.1)` calls before asserting on the "triggers currently running" / "watchers currently running" log lines. `_service_subprocess` returns as soon as any I/O arrives, so on a loaded CI runner that budget can be spent before the runner process has reached its status log. Example failure on an unrelated PR: https://github.com/apache/airflow/actions/runs/36232920075/job/108381612308 This polls until both lines have arrived (up to 300 iterations), mirroring the approach already used by `test_trigger_logger_fd_closed_when_removed`, and moves the supervisor `kill()` into a `finally` block so a failing assertion no longer leaks the runner process. Follow-up to #73397, which fixed the same class of flakiness in `test_trigger_lifecycle`. Verified locally: `test_trigger_log` passes on repeated runs; `prek run --from-ref upstream/main --stage pre-commit` passes. </details> --- ##### Was generative AI tooling used to co-author this PR? - [X] Yes — Claude Code (Fable 5.1) Generated-by: Claude Code (Fable 5.1) following [the guidelines](https://github.com/apache/airflow/blob/main/contributing-docs/05_pull_requests.rst#gen-ai-assisted-contributions) -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
