Skip to content

Free, Open Source, Self-Hosted WhatsApp API Gateway

fix(release): keep the staging version once promote has run

Yudhi Armyndharisb370e403aa4a
diff --git a/.github/workflows/release.yml b/.github/workflows/release.yml
index 3830b1a5c..61759f676 100644
--- a/.github/workflows/release.yml
+++ b/.github/workflows/release.yml
@@ -823,8 +823,9 @@ jobs:
# the staging tag would otherwise leak. On the happy path the release tags share the staging
# version, so the step below does not run at all (the deliberate retention documented above): the
# package-versions API is eventually consistent, and a read that does not list the new release
- # tags yet would count the staging tag alone and delete the version just published. On a failed
- # run the staging tag is alone and gets removed.
+ # tags yet would count the staging tag alone and delete the version just published. The same
+ # holds for a promote that failed after re-pointing some tags, so the step runs only when promote
+ # never ran, which is the one case where the staging tag is alone and gets removed.
cleanup-staging:
name: Clean up the staging tag
runs-on: ubuntu-latest
@@ -835,10 +836,10 @@ jobs:
needs: [docker, boot-smoke, image-scan, promote]
steps:
- name: Remove the staging tag (only when it is alone on its version)
- # Never after a successful promote: the release tags share this version by construction, so
- # a tag count of 1 there can only be a stale read. The count below still guards a promote
- # that failed after attaching some release tags.
- if: needs.promote.result != 'success'
+ # Only when promote never ran: once it has started re-pointing tags, the version can carry
+ # release tags that a stale read does not list yet. A failed or cancelled promote leaves the
+ # stray staging tag behind, the same retention as the happy path.
+ if: needs.promote.result == 'skipped'
# Best-effort: a failed tidy-up must not turn an otherwise successful release red.
continue-on-error: true
env:
diff --git a/src/common/release-staging-cleanup.spec.ts b/src/common/release-staging-cleanup.spec.ts
index 24599b73a..2dc43cbe8 100644
--- a/src/common/release-staging-cleanup.spec.ts
+++ b/src/common/release-staging-cleanup.spec.ts
@@ -6,8 +6,9 @@ import * as yaml from 'js-yaml';
* GHCR deletes by version, and after a successful promote the release tags share the staging tag's
* version. The cleanup job decides from a read of the package-versions API, which is eventually
* consistent: a read that does not list the new release tags yet shows the staging tag alone, and
- * deleting that version unpublishes X.Y.Z, X.Y and latest together. So the delete step must not
- * run at all once promote succeeded, whatever the read says.
+ * deleting that version unpublishes X.Y.Z, X.Y and latest together. A promote that failed partway
+ * may already have attached release tags too, and the same read cannot be trusted there. So the
+ * delete step runs only when promote never ran, whatever the read says.
*/
describe('release staging-tag cleanup', () => {
type Step = { if?: string; run?: string };
@@ -22,10 +23,10 @@ describe('release staging-tag cleanup', () => {
expect(job?.needs).toContain('promote');
});
- it('skips every delete once promote succeeded', () => {
+ it('skips every delete once promote ran at all', () => {
const deleting = (job?.steps ?? []).filter(step => /-X DELETE/.test(step.run ?? ''));
// Vacuity guard: a renamed job or a reworded command would leave nothing to check.
expect(deleting.length).toBeGreaterThan(0);
- for (const step of deleting) expect(step.if).toBe("needs.promote.result != 'success'");
+ for (const step of deleting) expect(step.if).toBe("needs.promote.result == 'skipped'");
});
});