The API Server should update its K8s Ready state to false when it looses connectivity to its required services

When the API server loses connectivity to is services (kv, kafka, etc) it should update the ready probe state appropriately so that the k8s will mark the pod as not ready. When connectivity is restored the state should go back to ready.

When the api server is in the "not ready" state the GRPC handler should return a not ready error and not process incoming requests