-
Bug
-
Resolution: Unresolved
-
Major
-
None
-
4.14
-
None
-
Hypershift Sprint 258, Hypershift Sprint 259, Hypershift Sprint 260, Hypershift Sprint 261, Hypershift Sprint 262
-
5
-
False
-
Description of problem:
After migrating the worker machine pools to new architecture (dedicated-v2), one of the Hosted Cluster's control plane pods including kube-apiserver pods are CrashLoopBackOff or in other error state.
After manually overriding the machine size from m5.xlarge to m5.4xlarge, the pods back to normal.
oc annotate hc/xxxx -n ocm-production-xxxx hypershift.openshift.io/cluster-size-override=m54xl
Version-Release number of selected component (if applicable):
4.14.30
How reproducible:
Not sure
Steps to Reproduce:
1.
2.
3.
Actual results:
The request serving nodes are undersized.
Expected results:
The request serving nodes are in right size and control plane pods are running.
Additional info:
#itn-2024-00139
