-
Bug
-
Resolution: Done-Errata
-
Major
-
None
-
4.14.z
Description of problem:
NodePool machine instances are failing to join a HostedCluster. The nodepool status reports InvalidConfig with the following condition:
- lastTransitionTime: "2024-05-02T15:08:58Z"
message: 'Failed to generate payload: error getting ignition payload: machine-config-server
configmap is out of date, waiting for update 5c59871d != 48a6b276'
observedGeneration: 1
reason: InvalidConfig
status: ""
type: ValidGeneratedPayload
Version-Release number of selected component (if applicable):
4.14.21 (HostedCluster), with HyperShift operator db9d81eb56e35b145bbbd878bbbcf742c9e75be2
How reproducible:
100%
Steps to Reproduce:
* Create a ROSA HCP cluster * Waiting for the nodepools to come up * Add an IdP * Delete ignition-server pods in the hostedcontrolplane namespace on the management cluster * Confirm nodepools complain about machine-config-server/token-secret hash mismatch * Scale down/up by deleting machine.cluster.x-k8s.io resources or otherwise
Actual results:
Nodes are not created
Expected results:
Node is created
Additional info:
The AWSMachine resources were created along with the corresponding ec2 instances. However, they were never ignited.
Deleting AWSMachine resources, resulted in successful ignition of new nodes.
Note: Please find logs from both HCP namespaces in the comment https://issues.redhat.com/browse/OCPBUGS-33377?focusedId=24690046&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-24690046
- is blocked by
-
HOSTEDCP-1677 Impact NodePool not bringing up nodes due to: configmap is out of date, waiting for update 5c59871d != 48a6b276
-
- Closed
-
- is cloned by
-
-
- New
-
- links to
-
RHEA-2024:3718
OpenShift Container Platform 4.17.z bug fix update
- mentioned on
