This went into upstream Kubernetes but I understand from https://issues.redhat.com/browse/OCPPLAN-4276 that we are not proactively taking upstream backports/patch release fixes.
I have seen this spamming kubelet logs on a number of 4.4.x OCP clusters in production so if it's easy to take the backport I'd suggest integrating it: https://github.com/kubernetes/kubernetes/pull/94432
It will prevent the logs from being spammed with the message "Setting volume ownership for %s and fsGroup set. If the volume has a lot of files then setting volume ownership could be slow, see https://github.com/kubernetes/kubernetes/issues/69699" across many volumes; the log entry will only land if it takes >30s.
Verified with: 4.6.0-0.nightly-2020-09-14-221526