Error message
Explanation
The Maia runner was told to shut down while it was running a component in your pipeline, and it was stopped before it could finish. This usually follows a deployment, a scale-in event, or a node replacement.Potential causes
- A deployment or rolling update replaced the Maia runner while pipelines were running.
- An autoscaler removed an instance that still had active work.
- The shutdown grace period on your platform is shorter than your longest-running component.
- A spot or preemptible node was reclaimed.
Resolve the error
- Compare the failure time with your deployment history and autoscaling events for the Maia runner.
-
Increase the termination grace period so it covers your longest-running component (for example,
terminationGracePeriodSecondsin Kubernetes orstopTimeoutin Amazon Elastic Container Service). - Schedule deployments and node maintenance outside the times your pipelines run, or pause schedules first.
- Avoid spot or preemptible nodes for Maia runners that run long jobs.
- Re-run the failed pipeline.
-
If you still need help, contact Maia Support and include the items listed below.
- The time of the shutdown event.
- What triggered the shutdown, if known.
Prevent future errors
To prevent this error occurring again:- Keep a minimum number of instances running during rolling updates.
- Configure scale-in protection while tasks are active, if your platform supports it.