-
Bug
-
Resolution: Done
-
Undefined
-
4.14.0
-
No
-
5
-
False
-
Description of problem:
A user destroying a HostedCluster can cause the HostedCluster to hang indefinitely if the destroy command times out during execution This is due to the hcp cli placing a finalizer on the HostedCluster during deletion which the cli tool later removes after waiting for some clean up actions to occur. If a user cancels the `hcp destroy cluster` command (or the command times out) while the cli is waiting for cleanup, then the HostedCluster will hang indefinitely with a DeletionTimestamp != nil. The cli tool should not be putting the HostedCluster into an un-reconcilable state. All this finializer cleanup logic belongs on the backend.
Version-Release number of selected component (if applicable):
4.14
How reproducible:
100%
Steps to Reproduce:
1. create an hcp cluster 2. destroy the hcp cluster with the cli tool and immediately abort the cli process 3.
Actual results:
HostedCluster is stuck indefinitely during deletion
Expected results:
HostedCluster is able to delete despite the cli being cancelled.
Additional info:
related to https://access.redhat.com/support/cases/#/case/03660218
- clones
-
OCPBUGS-23126 'hcp destroy' command can leave HostedCluster hanging indefinitely during cleanup
- Closed
- depends on
-
OCPBUGS-23126 'hcp destroy' command can leave HostedCluster hanging indefinitely during cleanup
- Closed
- links to