Is there a way to force a task manager to restart in case it fails to instantiate a metrics reporter? I just found that one of my task managers failed to report metrics back. After I checked the logging, I found that it failed to instantiate the dd metrics reporter due to a temporary network issue. However, in my case, I prefer the task manager to fail fast instead of continuing running. My app is running in a k8s env, so I was thinking to add a health check for the pods to ensure they report metrics. I was wondering whether there is better approach?