Handling un-cooperative pods during cordoning #147
Description
While in a perfect world no PDP would block the rescheduling of a pod due to a poorly written PDP because there would be a kyverno/OPA policy preventing them from existing...
Would it be possible to allow , through extra options in the CR, the ability to terminate pods that do not want to die (forceful: true, timeout: 300s) or something like that?
I really think this operator has a ton of potential to bring some operational visibility when cordoning nodes (usually one has no idea why they are cordoned- this operator provides a place to put comments)