In hosted mode, if configuration policies use pruneObjectBehavior and all the hosted (managed) clusters are deleted, the finalizers aren't cleaned up on the ConfigurationPolicy objects and the CRD gets stuck in a terminating state. This makes it so that ConfigurationPolicy objects can't be updated or created in the event that a new hosted (managed) cluster is added.
See the following issue for more context in the SD environment:
https://issues.redhat.com/browse/OHSS-17923
Likely relates to this PR not accounting for hosted mode:
https://github.com/open-cluster-management-io/config-policy-controller/pull/86
- relates to
-
ACM-3233 Hosted mode config-policy-controller uninstalls fail with pruneObjectBehavior
- Closed