-
Story
-
Resolution: Done
-
Normal
-
None
-
None
-
None
This will help us catch issues such as https://issues.redhat.com//browse/OCPBUGS-55428
Developer Story
As a developer working on the MCO repository, I want to add an informing WMCO end-to-end (e2e) test so that we can proactively monitor WMCO integration and catch regressions early.
Description
This story involves adding an informing-level WMCO (Windows Machine Config Operator) e2e test to the Machine Config Operator (MCO) repository.
The purpose is to integrate WMCO test coverage into MCO’s CI system with blocking merges (not optional)
This will help detect potential regressions related to Windows node handling as MCO evolves, while not introducing merge delays for unrelated changes.
Tasks include:
- Integrating the WMCO e2e test(s).
- Adding the informing test job into the relevant CI configurations (e.g., Prow).
Nice to have
- Add Slack or email alerting for failures of this test.
- Include a link in the MCO repo’s documentation explaining informing jobs and how to interpret their results.
Engineering Details
Follow the existing pattern in openshift-ovn-kubernetes and openshift-cluster-network-operator, where both incorporates Windows specific tests in their CI test suite
The windows-e2e-operator-test definition can be found in https://github.com/openshift/release/blob/master/ci-operator/step-registry/windows/e2e/operator/test/windows-e2e-operator-test-ref.yaml
For example, using openshift-ovn-kubernetes as a reference:
Test configuration:
Job configuration:
Acceptance Criteria
- A WMCO e2e test is added to the MCO repo and runs as an informing job in CI.
- The test reliably passes when there are no regressions, and failures are visible in PRs or test dashboards
- Stakeholders (MCO and WMCO teams) acknowledge successful integration.
- is caused by
-
WINC-1384 CI jobs failing due to kubelet CA test
-
- Closed
-
- links to