The GitHub Actions job "Tests" on airflow.git has succeeded. Run started by GitHub user droppoint (triggered by eladkal).
Head commit for run: ab73a04c53ae4129d9788f18ae75a247eafeac9a / Алексей Партилов <[email protected]> Fix race condition in KubernetesExecutor with concurrently running schedulers A race condition occurs in the _adopt_completed_pods function when schedulers are running concurrently. _adopt_completed_pods function doesn't keep track of which scheduler went down so it constantly tries to adopt completed pods from normally working schedulers. On Airflow setups with concurrently running schedulers and with a lot of short living DAG's it leads to race condition and open slots leak. You can find detailed analysis of this situation in GitHub issue here (apache#32928 (comment)). The _adopt_completed_pods function was refactored to the _delete_orphaned_completed_pods function, which removes only completed pods that are not bound to running schedulers. Co-authored-by: Vlad Pastushenko <[email protected]> Report URL: https://github.com/apache/airflow/actions/runs/7139880439 With regards, GitHub Actions via GitBox --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
