XML

Word

Printable

Epic Name:
improve_pre_merge_flake_detection
Activity Type:
Quality / Stability / Reliability
Story Points:
77
Blocked:
False
Blocked Reason:

Hide

None

Show
None
Ready:
False
Acceptance Criteria:
- Failure rate of flake detection job measured on merged PRs
- Infrastructure failures filtered from actual test failures
- Job timeout optimized with runtime metrics
- High-confidence flakes block PR merge
Component Fix Version(s):
None
Current Status:
Green
Epic Status:
To Do
Hierarchy Progress Bar:

100% To Do, 0% In Progress, 0% Done
Status Summary:

Hide

2026-02-16:
but do we still need to groom how can we improve the flakes in ci-health?...

Show
2026-02-16: but do we still need to groom how can we improve the flakes in ci-health?...

Goal

check-tests-for-flakes: Make the KubeVirt Pre-Merge Flake detection process voting job and also integrate flakes per sig into ci-health.

User Stories

Measure the failure rate of this job on merged PRs
Enhanced Filtering & Exclusions: Add more sophisticated filtering for infrastructure-related failures and Filter out failures during cluster initialization vs. actual test failures

Timeout & Resource Optimization: the 7h timeout seems very long. Adding metrics to track actual runtime vs. timeout

List of things not included in this epic, to alleviate any doubt raised during the grooming process.