Your web application is deployed on Google Kubernetes Engine (GKE), consisting of a backend service and a frontend service, both deployed as Kubernetes deployments. What is the best method to ensure uninterrupted communication between the frontend and backend services when pods are moved or restarted?