It would be great if we can define a health check for the container service. There might be cases where a process in the container or a dependent service might have issues and there is not way for us to direct to another container or take that container out.