[teeohhem]

Summary

E2E and the image-build jobs run on org-level larger runners (8-core-playwright, Large-Runner-x64-32, Large-Runner-ARM64-32, ubuntu-latest-arm64). Those don't move with the repo when it transfers to the ClickHouse org. If nothing in the new org has those labels, the jobs don't fail; they sit queued waiting for a runner.

This reads each label from a repo variable and falls back to today's label, so nothing changes until a variable is set. A value is either a plain label (ubuntu-24.04) or a JSON array of labels (["self-hosted","linux","x64"]), so the variables can also target self-hosted runners that are selected by a label set. On transfer day we can point the jobs at whatever runners ClickHouse has, or at standard GitHub runners, from Settings → Secrets and variables → Actions → Variables, without a PR.

Variable	Default	Used by
RUNNER_E2E	8-core-playwright	e2e-tests.yml (every PR)
RUNNER_LARGE_X64	Large-Runner-x64-32	Image builds in release.yml, release-nightly.yml
RUNNER_LARGE_ARM64	Large-Runner-ARM64-32	Image builds in release.yml, release-nightly.yml
RUNNER_ARM64	ubuntu-latest-arm64	arm64 build in release.yml, release-nightly.yml
E2E_WORKERS	unset (Playwright config default, 4 on CI)	e2e-tests.yml

E2E_WORKERS is already read by playwright.config.ts. It's passed through because 4 workers timed out on the 4-vCPU runner before; set it to 2 if RUNNER_E2E points at a smaller machine.

Standard-runner fallback: RUNNER_E2E=ubuntu-24.04 + E2E_WORKERS=2, RUNNER_LARGE_X64=ubuntu-24.04, RUNNER_LARGE_ARM64=ubuntu-24.04-arm, RUNNER_ARM64=ubuntu-24.04-arm. Image builds will be slower.

No changeset: CI only.

How to reproduce

1. On this branch with no variables set, the E2E jobs on this PR run on 8-core-playwright as before (check the "Set up job" log for the runner name).
2. Optional: set RUNNER_E2E=ubuntu-24.04 and E2E_WORKERS=2, re-run E2E, and confirm it picks up the standard runner and 2 workers. Unset both afterwards.

Not run: actionlint. I checked that the YAML parses and passes Prettier. vars is a documented context for both runs-on and strategy; the first CI run on this PR confirms the E2E path, and the release paths only run on main/nightly.

🤖 Generated with Claude Code

[changeset-bot]

⚠️ No Changeset found

Latest commit: 071e024

Merging this PR will not cause a version bump for any packages. If these changes should not result in a new version, you're good to go. If these changes should result in a version bump, you need to add a changeset.

This PR includes no changesets

When changesets are added to this PR, you'll see the packages that this PR includes changesets for and the associated semver types

Click here to learn what changesets are, and how to add one.

Click here if you're a maintainer who wants to add a changeset to this PR

[github-actions]

🔴 Tier 4 — Critical

Touches authentication, tenancy data models, the public API or shipped database config — or substantially changes the query rendering engine, background tasks, the OTel pipeline, image build, or release CI.

Why this tier:

• Background tasks or delivery pipeline substantially modified — 32 lines (bar: 30):
  • .github/workflows/release.yml

Review process: Deep review from a domain expert. Synchronous walkthrough may be required.
SLA: Schedule synchronous review within 2 business days.

Stats

• Production files changed: 0
• Production lines changed: 0
• Critical-path lines changed: 32
• Branch: claude/runner-label-vars
• Author: teeohhem

To override this classification, remove the review/tier-4 label and apply a different review/tier-* label. Manual overrides are preserved on subsequent pushes.

[github-actions]

PR Review

✅ No issues found.

[github-actions]

Duplicate code analysis

No new duplicate code introduced by this PR.

---

jscpd compared this branch against origin/main
across app, api and common-utils. Only new clones are listed; the codebase
currently holds 110 in total (1.22%).
Run yarn dupes locally to reproduce, or yarn dupes:all for the full picture.

[github-actions]

Deep Review

✅ No critical issues found.

This is a CI-only change making GitHub Actions runner labels configurable via repo variables with fallback to the current hardcoded labels. The ${{ startsWith(X, '[') && fromJSON(X) || X || 'default' }} idiom was verified correct on every normal path: unset/empty resolves to the hardcoded default, a plain label passes through, and a JSON array parses into a label set. &&/|| precedence, vars availability in strategy.matrix and runs-on, and the folded multi-line runs-on scalar are all valid. The E2E_WORKERS: ${{ vars.E2E_WORKERS }} passthrough sets an empty string when unset, which packages/app/playwright.config.ts:48 handles correctly (raw ? Number(raw) : NaN), so it falls back to the default workers.

🟡 P2 -- recommended

• .github/workflows/release.yml:810 -- The new matrix.runner/fromJSON runner expressions in release.yml and release-nightly.yml evaluate only on main/nightly triggers, so the override and JSON-array branches are unexercised by any PR-triggered run and a bad RUNNER_* value (e.g. malformed JSON) would first surface at release time.
  • Fix: Validate a configured RUNNER_* value through a workflow_dispatch smoke run before the first release that relies on it.
  • _{reliability, correctness}

🔵 P3 nitpicks (1)

• .github/workflows/e2e-tests.yml:14 -- The new RUNNER_E2E, RUNNER_LARGE_X64, RUNNER_LARGE_ARM64, RUNNER_ARM64, and E2E_WORKERS variables are discoverable only by reading inline comments across three workflow files, with no central listing of the variables and their expected format.
  • Fix: List the runner variables and their plain-label-or-JSON-array format in a CI/contributing doc so a maintainer can find the knobs on transfer day.

---

Reviewers (3): correctness, reliability, project-standards.

Testing gaps:

• Only the unset-fallback path runs in this PR's CI; the plain-label and JSON-array override branches are exercised only once the repo variables are set post-transfer.
• No PR-triggered workflow references the release.yml/release-nightly.yml jobs, so expression regressions there are caught only at release/nightly runtime.

_{No finding blocks merge automatically. A maintainer will expect P0/P1 findings in code this PR changes to be fixed; P2/P3 are your call -- fix or reply. Never fix findings about surrounding code here; reply instead. Do not widen the PR. How to respond}

[github-actions]

E2E Test Results

✅ All tests passed • 462 passed • 1 skipped • 1440s

Status	Count
✅ Passed	462
❌ Failed	0
⚠️ Flaky	2
⏭️ Skipped	1

Tests ran across 4 shards in parallel.

View full report →

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated
hyperdx-oss	Ready	Preview	Nov 5, 2026 9:27pm UTC
hyperdx-storybook	Ready	Preview	Nov 5, 2026 9:27pm UTC