DynamicResourceAllocation feature breaks TechPreview clusters

When installing a new cluster with TechPreviewNoUpgrade featureSet, Nodes never become Ready.

Logs from control-plane components indicate that a resource associated with the DynamicResourceAllocation feature can't be found:

E0804 15:48:51.094383       1 reflector.go:147] vendor/k8s.io/client-go/informers/factory.go:150: Failed to watch *v1alpha2.PodSchedulingContext: failed to list *v1alpha2.PodSchedulingContext: the server could not find the requested resource (get podschedulingcontexts.resource.k8s.io)

It turns out we either need to:

1. Enable the resource.k8s.io/v1alpha2=true API in kube-apiserver.
2. Or disable the DynamicResourceAllocation feature as TP.

For now I added a commit to invalidate this feature in o/k and disable all related tests. Please let me know once this is sorted out so that I can drop that commit from the rebase PR.

4.15

Always when installing a new cluster with TechPreviewNoUpgrade featureSet.

1. Install cluster with TechPreviewNoUpgrade featureSet (this can be done passing an install-config.yaml to the installer).
2. Check logs from one the control-plane components.

Nodes are NotReady and ClusterOperators Degraded.

Cluster is installed successfully.

Slack thread: https://redhat-internal.slack.com/archives/C05HQGU8TFF/p1691154653507499

How to enable an API in KAS: https://kubernetes.io/docs/tasks/administer-cluster/enable-disable-api/