The GitHub Actions job "Tests (AMD)" on 
airflow.git/k8s-executor-requeue-transient-5xx has failed.
Run started by GitHub user 1fanwang (triggered by 1fanwang).

Head commit for run:
52ea14370501e9935676d704b5eed1b8c0baccbc / 1fanwang <[email protected]>
Re-queue worker pods on transient 502/503/504 from the Kubernetes API

The KubernetesExecutor's shared pod-publish error handler re-queues
exceeded-quota / stale-version conflicts, 500, and 429, but fails the
task outright on 502/503/504 — the gateway / unavailable / timeout codes
that spike when the API server or an admission webhook is under load. A
shutting-down apiserver returns 503 with Retry-After on every graceful
roll, so an in-flight pod CREATE during a rollout currently kills the
task. Treat these as transient and re-queue them (bounded by
task_publish_max_retries), honoring Retry-After to back the loop off the
same way the 429 path does — matching what client-go already does for
all 5xx.

Report URL: https://github.com/apache/airflow/actions/runs/28190502399

With regards,
GitHub Actions via GitBox


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to