Description of problem:
We installed HPP CR with pvcTemplate based on CEPH storage class. CEPH got CriticallyFull and some hpp-pool pods couldn't create.

Version-Release number of selected component (if applicable):
4.12, 4.11, 4.10

How reproducible:
Only when there's a problem with the underlying storage class

Steps to Reproduce:
1. Create HPP CR with pvcTemplate based on CEPH
2. Use all CEPH storage
3. Delete HPP CR

Actual results:
$ oc get pods -A | grep hpp
openshift-cnv hpp-pool-29ab9406-85bc665cdb-wqz7j 1/1 Running 0 46h
openshift-cnv hpp-pool-4356e54b-7ccf5c44d-95tkr 1/1 Running 0 46h
openshift-cnv hpp-pool-7dfd761c-6ffd959c85-tfqds 0/1 ContainerCreating 0 19m

$ oc delete hostpathprovisioner hostpath-provisioner
hostpathprovisioner.hostpathprovisioner.kubevirt.io "hostpath-provisioner" deleted
(STUCK)

$ oc get jobs -n openshift-cnv
NAME COMPLETIONS DURATION AGE
cleanup-pool-4dd1b8bf 0/1 13m 13m
cleanup-pool-d1954b6a 0/1 13m 13m
cleanup-pool-edb68ab8 1/1 6s 13m

Expected results:
HPP CR deleted

Additional info:
As a W/A: delete the cleanup pods manually, they will be recreated and they will complete successfully.

HPP CR:
apiVersion: hostpathprovisioner.kubevirt.io/v1beta1
kind: HostPathProvisioner
metadata:
name: hostpath-provisioner
spec:
imagePullPolicy: IfNotPresent
storagePools:

• name: hpp-csi-local-basic
  path: "/var/hpp-csi-local-basic"
• name: hpp-csi-pvc-block
  pvcTemplate:
  volumeMode: Block
  storageClassName: ocs-storagecluster-ceph-rbd
  accessModes:
• ReadWriteOnce
  resources:
  requests:
  storage: 100Gi
  path: "/var/hpp-csi-pvc-block"
  workload:
  nodeSelector:
  kubernetes.io/os: linux