In https://issues.redhat.com/browse/OCPBUGS-24195 Lukasz is working on a solution to a problem both the auth and apiserver operators have where a large number of identical kube events can be emitted. The kube apiserver was granted an exception here, but the linked bug was never fixed.

These OpenShiftAPICheckFailed events are reportedly originating during bootstrap, and if bootstrap takes too long many can be emitted, which can trip a test that watches for this sort of thing.

Ideally the problem should be fixed and it sounds like Lukasz is on the path to one which we hope could be used for the apiserver operator as well. (start a controller monitoring the aggregated API only after the bootstrap is complete)

 Fix here would hopefully be to leverage what comes out of OCPBUGS-24195, apply it for the apiserver operator, and then remove the exception linked above in origin.