-
Bug
-
Resolution: Done-Errata
-
Major
-
None
-
4.17.0, 4.18.0
Description of problem:
The testing scenario is, IPI installation into shared VPC, with pre-configured firewall-rules, and the dns private zone, and the service account for control-plane nodes. According to the OCP doc "Required GCP permissions for shared VPC installation" (https://docs.redhat.com/en/documentation/openshift_container_platform/4.17/html/installing_on_gcp/installing-gcp-account#minimum-required-permissions-ipi-gcp-xpn_installing-gcp-account), as the control-plane service account had been specified, the permission "resourcemanager.projects.getIamPolicy" in the host project is not necessary. But, during "destroy cluster", the ".openshift_install.log" keeps telling below message which seems leading to "destroy cluster" hung. time="2024-12-03T13:33:57+08:00" level=debug msg="Service accounts: failed to fetch project IAM policy in project openshift-qe-shared-vpc: googleapi: Error 403: Permission 'resourcemanager.projects.getIamPolicy' denied on resource '//cloudresourcemanager.googleapis.com/projects/openshift-qe-shared-vpc' (or it may not exist).\nDetails:\n[\n {\n \"@type\": \"type.googleapis.com/google.rpc.ErrorInfo\",\n \"domain\": \"cloudresourcemanager.googleapis.com\",\n \"metadata\": {\n \"permission\": \"resourcemanager.projects.getIamPolicy\",\n \"resource\": \"projects/openshift-qe-shared-vpc\"\n },\n \"reason\": \"IAM_PERMISSION_DENIED\"\n }\n]\n, forbidden"
Version-Release number of selected component (if applicable):
4.18.0-0.nightly-multi-2024-12-02-195414
How reproducible:
Always
Steps to Reproduce:
0. supposing correct firewall-rules had been pre-configured within the shared VPC 1. activate the service account for the testing (see [1]) 2. "create install-config", and then insert the interested settings (see [2]) 3. pre-configure a dns private zone, binding to the shared VPC, and dns name matching the cluster's "<cluster name>.<base domain>." (see [3]) 4. "create cluster", and make sure it succeeds (see [4]) 5. "destroy cluster" (see [5])
Actual results:
"destroy cluster" stucks endlessly.
Expected results:
"destroy cluster" should finish within several minutes, along with all cluster resources getting deleted.
Additional info:
The .openshift_install.log is available at https://drive.google.com/file/d/1mpd474NmUkB9uB01y8ZIpWeJxmaz-zNO/view?usp=drive_link
- links to
-
RHEA-2024:11038 OpenShift Container Platform 4.19.z bug fix update