You assigned the bugzilla to apiserver-auth, but included kubelet and crio logs which we won't help us troubleshoot the apiserver. Please add must-gather from when the cluster was down.
Also, this BZ looks a lot like https://bugzilla.redhat.com/show_bug.cgi?id=1873816, have you checked it before opening this BZ?
This seems very similar to the known issue: https://github.com/kubernetes/kubernetes/issues/87615. As a workaround I would recommend restarting kubelet.
Also we have a known BZ https://github.com/kubernetes/kubernetes/pull/96549 where we are tracking this. The upstream fix https://github.com/kubernetes/kubernetes/pull/96549 is going to land in kube 1.20.
We will also need this patch: https://github.com/kubernetes/kubernetes/pull/95981.
Typo from c16: the BZ is https://bugzilla.redhat.com/show_bug.cgi?id=1873114.
Going to mark this BZ as a duplicate. We are working in the other BZ to get the fixes in. *** This bug has been marked as a duplicate of bug 1873114 ***