Repository navigation
ci(stream): add manual full measurement workflow - #159
Merged
Merged
Conversation
|
🦞👀 Pull request received. I will update this pull request when review starts. |
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
Add a manual-only, single-runner workflow for the existing full composed-stream benchmark. It uses Ubuntu 24.04, Node.js 24.22.1 and pnpm 11.28.4, installs from the frozen lockfile, and runs
pnpm checkbefore all 24 cells.The workflow validates sample counts, byte/lifecycle/accounting validity and source/build/toolchain identities. It retains bounded raw JSON for seven days, including parseable failure reports, with the commit/run/attempt in the artifact name. The guide explains dispatch identity, timeouts, failure handling and measurement limits.
Compatibility And Security
The workflow has only a manual trigger and read-only contents permission. Runtime code, public exports, dependencies, lockfiles and existing CI/release workflows are unchanged. It does not publish or deploy. A hosted run is a synthetic screen; this PR contains no performance results or queue-only, renderer, PTY or Rust benefit claims.
Validation
Passed strict YAML parsing with duplicate-key and alias rejection for all six workflows, Bash syntax checks, inline Node syntax, exact scope/privacy checks and pinned Oxfmt 0.75.0 standalone formatting/idempotence.
Independent complete-context P0–P3 review: passed in one Astra/high pass with no actionable findings.
Normal PR CI passed: all seven
pnpm checkjobs passed 193 tests each, with the named built public CLI smoke case visibly executed in every job. Browser smoke and workflow lint also passed. Both CodeQL analyses passed. Local project execution remained unavailable because the task had no qualified dependency installation.The full 24-cell workflow has not been dispatched. After landing, one separately verified exact-commit branch dispatch will supply the raw measurement report; source and CI validation alone are not measurement evidence.
Workflow failure handling and existing benchmark validation considered
Browser, Node.js, and Worker impact considered
CHANGELOG.mdupdated for the operational changeNo credentials, private paths, private hosts, or sensitive logs included