How to use latency of a service deployed on Kubernetes to Scale the deployment? How to use latency of a service deployed on Kubernetes to Scale the deployment? kubernetes kubernetes