Repository navigation
fix(e2e): pass the product version to chart-e2e, pin the linter version - #265
Merged
Merged
Conversation
golangci-lint-action was invoked without a `version:` input, so CI always installed the newest release. That makes the pipeline fail on an upstream release with no change here, and it silently diverges from the v2.12.1 the Makefile already pinned. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
chainsaw-e2e forwarded --set product_version, but chart-e2e did not, so the chart-based run always used the operator default no matter what the CI matrix selected. The product-version dimension was real for one job and decorative for the other. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
publish.yml could only be triggered by a push to main, so the credentials it needs could not be verified on demand. An expired HELM_CHARTS_REPO_TOKEN went unnoticed for weeks because nothing happened to land on main during that window, and it only surfaced when an unrelated change did. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
The chart CRDs were generated by an older controller-gen than config/crd/bases. Only config/ is covered by the check-crds-sync job, so this copy could drift indefinitely without CI noticing. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Why
Part of the 0.4.0 release baseline alignment.
chart-e2e ignored the product-version matrix.
chainsaw-e2eforwarded--set product_version, butchart-e2edid not — so the chart-based run alwaysused the operator default no matter which version the CI matrix selected. The
matrix dimension was real for one job and decorative for the other, which is
worse than having no matrix: the job reports coverage it never had.
golangci-lint was floating.
golangci-lint-actionwas invoked with noversion:input, so CI always installed the newest release. That breaks thepipeline on an upstream release with no change here, and it diverges from the
v2.12.1 the Makefile already pinned.
The chart CRDs had drifted from
config/crd/bases.check-crds-synconlycovers
config/, so that copy can drift indefinitely without CI noticing.publish.yml could only be triggered by a push to main, so its credentials
could not be verified on demand. That is how an expired
HELM_CHARTS_REPO_TOKENwent unnoticed for weeks.
Verification
make lint(v2.12.1)make testcheck-crds-sync(simulated per the CI logic, clean tree)make helm-crd-syncworkflow_dispatchparses as expectedThe
--setfix itself is exercised by this PR's ownchart-e2ejob, which isthe first run where the matrix value actually reaches the CR.
🤖 Generated with Claude Code