Clarify documentation for Kubenetes Service configuration

[TimMoore]

Right now the documentation for routing external traffic to Akka in a Kubernetes environment contains this:

For production traffic e.g. HTTP use a regular service or an alternative ingress mechanism. With an appropriate readiness check this results in traffic not being routed until bootstrap has finished.

apiVersion: v1
kind: Service
metadata:
  labels:
    app: akka-bootstrap-demo
  name: "integration-test-kubernetes-dns"
spec:
  ports:
  - name: http
    port: 8080
    protocol: TCP
    targetPort: 8080
  selector:
    app: "akka-bootstrap-demo"
  type: LoadBalancer

This will result in a ClusterIP being created and only added to Endpoints when the pods are ready

Note that the app is the same for both services as they both refer to the same pods.

A lot of readers who aren't familiar with the details of Kubernetes services might not be able to interpret this and are likely to copy the example without looking into alternatives. Indeed we seem to have copied it uncritically into some of our own example projects and guides. I think our default recommendation should be to use the ClusterIP service type (which is the default if you don't specify type) and then use Ingress to get traffic into the cluster. We might want to point readers to other resources to help them decide what to do for their own production services, such as one of these:

• https://medium.com/google-cloud/kubernetes-nodeport-vs-loadbalancer-vs-ingress-when-should-i-use-what-922f010849e0
• https://matthewpalmer.net/kubernetes-app-developer/articles/kubernetes-ingress-guide-nginx-example.html

(The last sentence is only relevant for the DNS bootstrap method, so that should probably move to that page.)

[chbatey]

Good idea!