[prometheus-node-exporter] Prevent node exporter from being scheduled on fargate

[mariuskimmina]

Is your feature request related to a problem ?

When running on EKS clusters with a mix of fargate and EC2 nodes the node exporter pods can end up unschedulable and stuck in a pending state.

Describe the solution you'd like.

Add a nodeAffinity to the node-exporter by default

  affinity:
    nodeAffinity:
      requiredDuringSchedulingIgnoredDuringExecution:
        nodeSelectorTerms:
        - matchExpressions:
          - key: eks.amazonaws.com/compute-type
            operator: NotIn
            values:
            - fargate

Describe alternatives you've considered.

None

Additional context.

There have been many previous Issues about this:

#2876
#2908
#925

Grafanas k8s-monitoring charts have also added a node affinity to the node-exporter by default which prevents it from running on fargate - see grafana/k8s-monitoring-helm#362.
I think it would be useful to do the same here.

If my understanding is correct, then for people that have already added this nodeAffinity this would not be a breaking change, it would simply be redundant - but this I am not completely sure about.