Seems like we can't use oc edit or apply to and edit an existing CR, because the existing deployment will not react to the change. Probably a problem with the reconciler.    

    4.17.0

Only after all reconcilliation and the operator is in a stable state, will CR in-place changes not work. If the operator is in an unstable state (e.g., not all pods have rolled out), then CR changes will seem to work (but really its just reconciling over and over again since the operator is trying to move into a READY state).

    1. Installed 4.17 CRO-operator
    2. Create a CRO CR like this:
apiVersion: operator.autoscaling.openshift.io/v1
kind: ClusterResourceOverride
metadata:
  name: cluster
spec:
  podResourceOverride:
    spec:
      memoryRequestToLimitPercent: 50
      cpuRequestToLimitPercent: 25
      limitCPUToMemoryPercent: 200
  deploymentOverrides:
    replicas: 1
    nodeSelector:
      node-role.kubernetes.io/worker: ""
    tolerations:
    - key: "key"
      operator: "Equal"
      value: "value"
      effect: "NoSchedule"
3. Edit it with oc edit and change replicas to 2 (alternatively use oc apply with an existing yaml definition of 2 replicas)
4. oc get deployment clusterresourceoverride -n clusterresourceoverride
5. Notice replicas do not change.

    Replicas do not change and stays at 1 pod.

    Replicas should scale to 2 pods.