Use the v2beta2 syntax for CPU and Memory autoscaling

[gary-archer]

I made a small change so that the syntax output for the HPA is now correct for v2beta2 for CPU and memory:

metrics:
- type: Resource
  resource:
    name: cpu
    target:
      type: Utilization
      averageUtilization: 80

Customers then need to install the metrics-server, which could be done like this in a basic development install:

kubectl apply -f https://github.com/kubernetes-sigs/metrics-server/releases/latest/download/components.yaml --kubelet-insecure-tls

By default I then get this issue, and from googling it seems that resource limits should also be activated for this feature to work. Eg also set this in the Helm chart:

resources:
  limits:
    cpu: 500m
  requests:
    cpu: 500m

For now I will add some brief notes to the autoscaling tutorial to indicate that you may also need to set CPU / memory resource limits. This behaviour could change in future I think.