fix: CI workflow permissions and E2E secret cleanup#56
Conversation
|
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: Organization UI Review profile: CHILL Plan: Pro Run ID: 📒 Files selected for processing (2)
📝 WalkthroughWalkthroughTwo files were modified: the GitHub Actions workflow now explicitly grants package write permissions to the release job, and the integration test cleanup function adds a blocking wait for CNPG-managed Kubernetes secrets to be completely deleted after Helm uninstall. Changes
Poem
🚥 Pre-merge checks | ✅ 2 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (2 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches📝 Generate docstrings
🧪 Generate unit tests (beta)
Comment |
Up to standards ✅🟢 Issues
|
| Metric | Results |
|---|---|
| Duplication | 0 |
TIP This summary will be updated as you push new changes. Give us feedback
Summary
Grant
contents: writeandpackages: writeto thereleasejob intag-release.yaml.When calling
release.yamlviaworkflow_call, the called workflow's permissionsare capped by the caller. The
releasejob had no explicit permissions, so itdefaulted to the
pull_requesttrigger's read-only scope, failing validation evenwhen the job was skipped by its
ifcondition.Wait for CNPG-managed secrets to be garbage collected between E2E tests.
TestNodesAddNodeZeroDowntimewas timing out on every open PR becauseuninstallChart()only waited for clusters, pods, jobs, and PVCs — not theoperator-generated secrets (
<cluster>-ca,<cluster>-server). These are ownedby the Cluster resource and GC'd asynchronously. When the next test reused the
same cluster names, the new CA didn't match the stale server cert, causing a
permanent
x509: ECDSA verification failurebetween the operator and instance.