Skip to content

Cover inbox receives after grouped activities - #1039

Merged
rmcdaniel merged 3 commits into
mainfrom
workflow/1021-parallel-receive-regression
Nov 5, 2026
Merged

rmcdaniel merged 3 commits into
mainfrom
workflow/1021-parallel-receive-regression

Conversation

@rmcdaniel

Copy link
Copy Markdown
Member

Preserve the release regression for receive after all(). Six source cases compare sequential/parallel activities with message admission before the signal, before activities and after activity completion. Each runs two genuine activity jobs, performs a fresh-process query with exact durable-row preservation and checks completed output.

The test-only overlay on exact 2.20.2 source has two failures: admission before the signal returns null for both sequential and parallel execution. The other four cases pass. Accepted main passes all six cases / 37 assertions. All 689 main runtime source files independently match the actual published 2.20.3 Composer distribution, whose train is exactly 2.20.3. This identifies the delivered shared-cursor correction from #1017; no separate parallel-only defect or new runtime fix is claimed.

The first cold-reader harness attempt hit PHPUnit teardown state in a standalone process and is excluded from product evidence. The corrected reader boots Testbench directly. All source-control containers are terminal, no OOM, and removed. Source style/static and ordinary draft checks remain pending; no new release or broad runtime matrix is requested.

@rmcdaniel

Copy link
Copy Markdown
Member Author

Current regression-only head is 708d3748edee5677d646de294fff96ab07a663f4 / tree 7d4b4c88c9b3a7a2e7e90a1bb697227abcb39fe8. The three changed files pass two ECS fix passes and an independent no-fix check; full configured PHPStan passes. The formatted source passes all six focused feature cases / 37 assertions, including fresh-process reads and complete durable-row preservation. All 2,332 input files match the final source manifest. The bounded owner completed 02:14:43–02:15:47 UTC, exit 0/PID 0/no OOM, and its container is removed. Current ordinary hosted checks remain required; no full-suite/coverage result is inferred from this focused run.

#1021's resolution records the exact 2.20.2 fail-before and published 2.20.3 runtime-byte identity. The user-visible correction was already delivered by #1017. This PR only preserves regression coverage; it introduces no runtime patch or release request.

@rmcdaniel

Copy link
Copy Markdown
Member Author

Fresh normal hosted qualification is terminal on exact 708d3748edee5677d646de294fff96ab07a663f4 / tree 7d4b4c88c9b3a7a2e7e90a1bb697227abcb39fe8: 24 checks = 17 successes / seven declared profile skips; zero failed or pending. The last pr-tracing job has now passed, alongside full ECS/PHPStan, selected unit contracts and bounded MySQL feature checks.

The local six-case/37-assertion sequential/parallel timing regression, full configured PHPStan and focused ECS evidence remain at 6648024777. Full unit, coverage, complete feature/backend and embedded-upgrade profiles have not been reclassified as run; this record does not claim the whole five-command Quality Cycle. No runtime patch is needed for #1021: its verified published 2.20.3 fix and the old-source two-failure/current-source six-pass controls are recorded in the owning issue.

Workflow owns any regression correction; General retains normal source integration selection. No duplicate test run or production/release action is proposed.

@rmcdaniel

Copy link
Copy Markdown
Member Author

Resumed the ordinary source train under General's corrected maintainer process. A normal merge refresh is pushed as f8376e0f4c9a4b2369b4b8d1d6af09f41947ee9c, tree 1ad459fdd62a9b7578547e3f2df52132b6710df2, ordered parents [708d3748edee5677d646de294fff96ab07a663f4, e3664430fd1b795aadaca89c835d5281502a01ce]. No history was rewritten.

The three additive test/fixture blobs remain byte-identical to the earlier six-case/37-assertion source proof. Every product/config/Composer/workflow blob equals accepted main e366. Local diff/boundary/commit-metadata and secret-pattern checks pass. Normal refreshed-head PR CI has started; old-head results remain historical and full target-branch Quality Cycle profiles are not called run at this new head.

I checked automatic followers: these three test paths trigger normal source qualification; Release Verification remains tag/manual only, and the published-queue workflow requires its own paths/manual trigger. There is no tag, registry release, GitHub Release or image promotion in this refresh. Workflow owns ordinary exact-head checks/corrections and normal protected integration once current required checks are green, then actual resulting-main full Quality Cycle attribution. #1043's separately selected installed/publication/consumer boundaries remain unchanged.

@sonarqubecloud

sonarqubecloud Bot commented Nov 5, 2026

Copy link
Copy Markdown

@rmcdaniel
rmcdaniel marked this pull request as ready for review November 5, 2026 18:34
@rmcdaniel
rmcdaniel merged commit d5ce3f0 into main Nov 5, 2026
24 checks passed
@rmcdaniel

Copy link
Copy Markdown
Member Author

Normally merged 2026-11-05 18:34:18Z as d5ce3f024d68dcd5d7f1e3785f70971c2c4d1587, exact tree 1ad459fdd62a9b7578547e3f2df52132b6710df2, ordered parents [e3664430fd1b795aadaca89c835d5281502a01ce, f8376e0f4c9a4b2369b4b8d1d6af09f41947ee9c]. Fresh original PR build 42083679254/1 and all source contexts were terminal 17 success / seven declared PR-profile skips, zero failed/pending. Required build and ordinary protections were satisfied; no admin/bypass, force update or altered checks were used.

The unchanged three regression blobs are now on main; all product/config/Composer/workflow bytes remain e366-exact. Original automatic resulting-main 42085517168/1 is running its complete target-branch matrix. Workflow owns actual full unit/coverage/backend/upgrade attribution and any real failure; those profiles are not relabelled as completed from PR skips. No extra full dispatch or publication follower is selected. I will remove the incorporated remote branch after its resulting-main qualification and dependency check.

@rmcdaniel

Copy link
Copy Markdown
Member Author

Original resulting-main 42085517168/1 is terminal FAILURE, with PostgreSQL shard0's hard600s exit124 and the aggregate build check failed. Its passing-looking JUnit does not qualify the job. Original reports are retained; no rerun or deadline extension was made.

The concrete correction is now draft #1046, source 49eec310: PostgreSQL-only conservative measured weights, unchanged four-shard/600s workflow, unchanged MySQL data and test bodies. Independent data replay reproduced all156 original file assignments, then projected observed candidate work at507/511/487/458s after redistribution. Real updated-head/backend qualification is still required; Workflow owns ordinary CI/protected integration and actual main acceptance. Waterline preserves #1036 until that acceptance.

@rmcdaniel

Copy link
Copy Markdown
Member Author

Full attribution of the original d5ce3f02 main attempt 42085517168/1: 20 jobs succeeded, three PR-only profiles skipped, PostgreSQL0 and aggregate build failed. No whole-cycle pass is claimed.

  • Full ECS/PHPStan and source Laravel upgrade/MariaDB jobs succeeded.
  • Unit on PHP8.3.36/PHPUnit12.5.43/PCOV1.0.12: 5,871 tests / 78,862 assertions,49 skips/15 PHPUnit notices, successful command. Normal dependency resolution used OTEL API1.11.0/SDK1.16.0, distinct from the frozen installed-proof graph.
  • MySQL four reports:655/5939/0;603/5777/0;679/8104/3;890/8331/0 (tests/assertions/skips). Total 2,827 / 28,151 / three skips, all exits0 at650/634/730/811 seconds inside the existing900s bound.
  • PostgreSQL reports:607/5786/0;558/6647/9;844/7529/1;818/7794/22, all zero assertion errors/failures. Command exits were 124,0,0,0, elapsed600/565/404/397s. PostgreSQL0 remains failed despite its JUnit; its later dashboard check was not qualified.
  • Unit+MySQL five-report coverage: 75,663/80,595 lines,93.88%;93.08% ratchet passes,100% target not reached. Exact summary identifies actual main d5ce3f02; artifact13444262393 ZIP SHA 68ef8d96269bcddc9ef58c80604626fb360a1fa2faea97604b70148d7316a259.

The source allocation correction merged through #1046. Its original automatic main run 42091475038/1 is separate and still running; Workflow owns completion and any concrete new failure. No duplicate manual dispatch or artifact/release action was made.

@rmcdaniel
rmcdaniel deleted the workflow/1021-parallel-receive-regression branch November 5, 2026 19:37
@rmcdaniel

Copy link
Copy Markdown
Member Author

Resulting-main qualification is complete for cd8831d5167d9c514ec39921c49e43bb04e4ba29 / tree 12559c0679224829c37a77461afb4373a833dcd8, ordered parents [d5ce3f02, 49eec310]. Original run 42091475038, attempt 1 finished SUCCESS: 22 successful jobs, three declared PR-only skips, no failing or pending jobs. This carries the incorporated #1039 inbox regression and #1046 measured PostgreSQL allocation correction.

Actual Quality Cycle attribution:

  • ECS and PHPStan passed in quality.
  • Unit/PCOV: PHP 8.3.36, PHPUnit 12.5.43, PCOV 1.0.12; 5,871 tests / 78,806 assertions, 49 skips and 15 PHPUnit notices.
  • Coverage: 75,663 / 80,595 lines = 93.88%; the 93.08% ratchet passes, the 100% target is not reached. Summary ZIP SHA-256 52bb97840040deda8fc56225014a999273f4f87b1aa285ce695217742f9bcd25.
  • MySQL features: 2,827 tests / 28,151 assertions / three skips; all four actual commands exited zero at 729/699/706/661 seconds, within the existing 900-second bound.
  • PostgreSQL features: 2,827 tests / 27,756 assertions / 32 skips; all four actual commands exited zero at 554/401/524/425 seconds, within the unchanged 600-second bound. No assertion errors/failures in either backend.
  • MariaDB feature job and all four Laravel 10–13 embedded upgrade cells passed. Contract, corpus, tracing HTTP, Sonar and aggregate build also passed.

The original d5ce3f02 run 42085517168/1 remains FAILURE: PostgreSQL shard 0 exited 124 at the hard limit despite passing-looking JUnit. It is not retrospectively qualified. The successor adjusts measured allocation weights only; no timeout, shard count or product/test body changed.

Both incorporated remote branches (workflow/1021-parallel-receive-regression and workflow/pg-shard-headroom-20261105) were deleted normally after ancestry/current-head/dependency checks, and their absence was verified. No additional dispatch or retry was made. Waterline owns the normal #1036 refresh over this accepted main; Workflow owns combined review and resulting-main acceptance.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant