-
Task
-
Resolution: Unresolved
-
Undefined
-
None
-
None
-
None
-
None
-
False
-
-
False
-
None
-
None
-
None
-
None
Test: OCP-39030 - Re-queue on Windows machines edge cases
Platforms: Azure, GCP, vSphere
Versions: 4.21, 4.22
Total Failures: 4 in last 7 days
File: test/extended/winc/winc.go
Root Cause Analysis (per platform)
Azure: getWindowsMachineSetName() in utils.go only searches for "winworker" substring, but Azure MachineSet is named "windows".
Error: MachineSet with substring winworker not found in Prow's cluster
Fix: PR #29179 - adds defaultWindowsMS ("windows") matching.
Status: Fixed
GCP: Timeout - Windows nodes not ready within 20 minutes after WMCO scale down/up.
Error: Windows nodes are not ready after waiting up to 20m0s minutes
Status: Intermittent timing issue, needs investigation
vSphere: dockercfg secret timeout - same root cause as WINC-1578.
Error: timed out waiting for the condition (client.go:424)
Status: Blocked by WINC-1578