Skip to content

chore(ci): drop redundant nightly-release-tag-v5-next workflow#24751

Open
AztecBot wants to merge 1 commit into
nextfrom
cb/drop-redundant-v5-next-nightly-tagger
Open

chore(ci): drop redundant nightly-release-tag-v5-next workflow#24751
AztecBot wants to merge 1 commit into
nextfrom
cb/drop-redundant-v5-next-nightly-tagger

Conversation

@AztecBot

Copy link
Copy Markdown
Collaborator

What

Deletes .github/workflows/nightly-release-tag-v5-next.yml, keeping the pre-existing nightly-release-tag.yml as the single nightly tagger.

Why

The two scheduled workflows had started pushing the identical nightly tag on the public repo, so the second run of the night failed on a duplicate tag:

  • nightly-release-tag.yml (4 AM UTC) — matrix tagger: public → v5-next, private → next + v5-next.
  • nightly-release-tag-v5-next.yml (5 AM UTC, added in 7ac407c) — always tags v5-next on the public repo only.

Both check out v5-next and build v${manifest.version}-nightly.$(date +%Y%m%d), so on the public repo they produce the exact same tag. On the night of 2026-07-15 the 4 AM run (#497) created v5.0.0-nightly.20260715, and the 5 AM run (first-ever scheduled run) failed:

 ! [rejected]  v5.0.0-nightly.20260715 -> v5.0.0-nightly.20260715 (already exists)
error: failed to push some refs to '.../aztec-packages'
hint: Updates were rejected because the tag already exists in the remote.

7ac407c added the -v5-next.yml tagger on the assumption that nightly-release-tag.yml tagged next on public. Two days later #23984 repointed that workflow's public matrix leg from [next, v5-next] to [v5-next] ("public should not produce a next nightly tag; v6/next nightly lives on private"), which made the two workflows' public output identical.

Why drop this one (the smaller of the two)

nightly-release-tag.yml is a strict superset:

nightly-release-tag.yml (kept) nightly-release-tag-v5-next.yml (removed)
public (schedule) tags v5-next tags v5-next (duplicate)
private (schedule) tags next + v5-next does not run (if gate is public-only)
manual workflow_dispatch yes (matrix) yes (v5-next)

Removing the smaller file eliminates the duplicate-tag failure while preserving every current behaviour: the surviving workflow already tags public v5-next, private next (v6) + v5-next (v5), and keeps a workflow_dispatch button for ad-hoc runs. Deleting the other file instead would have silently dropped all private-repo nightly tagging.

Testing

Workflow-file-only change; no code paths affected. Verified the surviving nightly-release-tag.yml at the next tip still carries the full repo-dependent matrix (private → [next, v5-next], public → [v5-next]) and workflow_dispatch.

Full diagnosis: https://gist.github.com/AztecBot/9aa164d7949a3cf68c7aebaaa1333649


Created by claudebox · group: slackbot

@AztecBot AztecBot added ci-draft Run CI on draft PRs. ci-no-fail-fast Sets NO_FAIL_FAST in the CI so the run is not aborted on the first failure claudebox Owned by claudebox. it can push to this PR. labels Jul 16, 2026
@Thunkar
Thunkar marked this pull request as ready for review July 16, 2026 14:45
@Thunkar
Thunkar requested a review from charlielye as a code owner July 16, 2026 14:45
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

ci-draft Run CI on draft PRs. ci-no-fail-fast Sets NO_FAIL_FAST in the CI so the run is not aborted on the first failure claudebox Owned by claudebox. it can push to this PR.

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants