Your noProxy configuration does not seem to include the cluster defaults which you'll need to do until https://github.com/openshift/cluster-network-operator/pull/245 is merged and https://bugzilla.redhat.com/show_bug.cgi?id=1738284 is closed.
Removing the dependency so that CI allows to push both PRs in parallel
The linked PRs all merged, moving to MODIFIED.
Verify this bug with 4.2.0-0.nightly-2019-08-15-205330 kube-apiserver and openshift-apiserver pods are running well.
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHBA-2019:2922