In the OpenShift conformance test suite the apiserver-incluster-availability monitor always uses the default 6443 port to connect to the kube-apiserver. This will cause all requests to time out if the kube-apiserver is configured to use a non-default port (like on hypershift).

    4.19 and above

    always

    1. Deploy a 4.19 hypershift cluster with a non-default kube-apiserver port.
    2. Run the OpenShift conformance testsuite against this cluster.
    

    The monitor uses the port 6443 port the connect to the kube-apiserver

    The monitor should use the port specified in the infra.Status.APIServerInternalURL when connecting to the kube-apiserver

    The port gets discarded here: https://github.com/openshift/origin/blob/85efd761c5387e6b4ada48eeb0d10bd2858b1bc7/pkg/monitortests/kubeapiserver/disruptioninclusterapiserver/monitortest.go#L361