-
Bug
-
Resolution: Done-Errata
-
Critical
-
CNV v4.15.0
-
None
-
5
-
False
-
-
False
-
aaq-operator-rhel9-container-v4.16.0-72
-
---
-
---
-
-
CNV Virtualization Sprint 251
-
No
Description of problem:
It allows to create multiple AAQ objects and the new one is in error state (which is probably fine), but when I tried to remove second AAQ objects - aaq-operator pod becomes in "CrashLoopBackOff" state and second AAQ can be removed only when I edit it and remove the finalizer manually # new AAQ in error state: $ oc get aaq NAME AGE PHASE aaq-kubevirt-hyperconverged 2m29s Deployed test-aaq 38s Error $ oc describe aaq test-aaq | grep Event -A 5 Events: Type Reason Age From Message ---- ------ ---- ---- ------- Warning ConfigError 18s (x9 over 2m11s) operator-controller Reconciling to error state, unwanted AAQ object # But can't delete it: test-aaq 3m41s Deleting # and when tried to delete - aaq-operator pod become in CrashLoopBackOff state: aaq-operator-669fbf7455-hwm5c 0/1 CrashLoopBackOff 5 (2m49s ago) 3m30s
Version-Release number of selected component (if applicable):
4.15
How reproducible:
100%
Steps to Reproduce:
1. enable AAQ 2. create second AAQ object 3. delete second AAQ
Actual results:
aaq-operator pod is in CrashLoopBackOff state (until the second AAQ object deleted by manual removing of finalizers)
Expected results:
Not sure but probably we should prevent from creating of multiple AAQ objects? (same as we do for Kubevirt) $ oc create -f new-kubevirt.yml Error from server: error when creating "new-kubevirt.yml": admission webhook "kubevirt-create-validator.kubevirt.io" denied the request: Kubevirt is already created OR have aaq-operator POD in a good shape all the time
Additional info:
- links to
-
RHEA-2023:122979 OpenShift Virtualization 4.16.0 Images