K8 Operator task is marked as failed without errors but container still running

[kand617]

Airflow version: 2.6.3

We are using K8 Pod Operator and we often see that the log randomly stops. No Error or anything.

It ends up retrying and seeing the pod is still active and reattaches if retry count is within the limit.
I suspect the worker monitoring the pod dies and does not reattach properly.

As the container is not terminated it should keep monitoring the logs.

[kand617]

bump +