Hello! We are running Prefect 2 with a Kubernetes agent on the Azure Kubernetes service having recently upgraded from v1. We are now finding that our agent seems to become marked as unhealthy in the cloud UI after some time and no longer picks up jobs. Despite this, the agent pod seems to still be running without any issues.

We're having the exact same issue at the moment but with AWS Kubernetes. Restarting the container seems to get it back to healthy again but it doesn't last

same here, but we’re using ECS Fargate to run the agent