We have 40+ micro services running in pods. Each service have 3 versions so we have 3 pods per service. When we started we had 1 ingress gateway which load balance the incoming HTTP traffic into all of these 40+ backend services. We had been experiencing 500 errors and could not connect to the services. Only was around this is if we point the API endpoint to the Pod IP and not the Ingress external IP. This then disables the load balancing between the versions. Is it right that we have only 1 ingress gateway or should we group pods in. multiple ingress gateway. We have 3 environments (Stage, Perf and Production). In each environment is only 1 ingress gateway. Please advise. Thank You.