Repository navigation
test(bake): wait for the held compile stamp, never sample HeldCount (red on #6132 shard 5) - #6169
Merged
Merged
Conversation
AGatedSweepHearsItsOwnCompileTest.AProvisionalSweep_ReportsCompiled_ThoughItsStampIsHeld read MeshPublicationGate.HeldCount once, right after the warm answered Compiled. By design the process-local witness (LocalNodeTypeBuilds) is published BEFORE the stamp, and the stamp is offered to the gate only after the release-create observation completes — so the read could land in between and see 0 (core #6132 shard 5, run 37338956752: 'Expected 0 to be greater than 0 because the compile stamp waits for the bake's verdict'; #6132 touches none of this). The test now waits for HeldCount > 0 within TestTimeouts.Convergence, then checks the record does not carry the build. Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Contributor
Test Results (shard 4)0 tests 0 ✅ 0s ⏱️ Results for commit 9f28073. |
Contributor
Test Results (shard 5)0 tests 0 ✅ 0s ⏱️ Results for commit 9f28073. |
Contributor
Contributor
Test Results (shard 3)0 tests 0 ✅ 0s ⏱️ Results for commit 9f28073. |
Contributor
Test Results (shard 0) 1 files - 2 1 suites - 2 2m 17s ⏱️ - 1m 7s Results for commit 9f28073. ± Comparison against base commit 64a828a. This pull request removes 566 and adds 598 tests. Note that renamed tests count towards both. |
Contributor
Test Results (shard 1)510 tests - 1 243 510 ✅ - 1 241 6m 19s ⏱️ +37s Results for commit 9f28073. ± Comparison against base commit 64a828a. This pull request removes 1753 and adds 510 tests. Note that renamed tests count towards both. |
Contributor
Test Results 2 files 2 suites 8m 36s ⏱️ Results for commit 9f28073. |
systemorph-com
Bot
added
the
thread:pr-systemorph-meshweaver-6169-d2a470
This was referenced Oct 8, 2026
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
1 participant
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.
Red
Core #6132 shard 5 (run 37338956752) failed in
MeshWeaver.Hosting.Test.AGatedSweepHearsItsOwnCompileTest.AProvisionalSweep_ReportsCompiled_ThoughItsStampIsHeldwithExpected 0 to be greater than 0 because the compile stamp waits for the bake's verdict. #6132 (instance reboot) touches nothing on this path: not the publication gate, the pre-warmer or the compile pipeline. So this is a race in the test that can hit any PR.Root cause
The test reads
MeshPublicationGate.HeldCountonce, straight afterWarmOneanswersCompiled. The two events it compares are ordered the other way round on purpose:DynamicTypePreWarmer.WarmOneanswers from the process-local witness (LocalNodeTypeBuilds), which "is published BEFORE the (possibly held) stamp" (Broken NodeType BinaryClickerV2/BinaryToggle never bakes, failing nodetype_bake readiness and stalling portal rollouts #5544).admission.Publish("NodeType compile stamp for …")inNodeTypeCompilationHelpers) only inside the release-create observation's subscription, which comes later.So the single read can land between the two and see 0.
Fix (test only)
Wait for
HeldCount > 0withinTestTimeouts.Convergenceinstead of sampling it, using the sanctioned interval re-query. Then check that the shared record does not carry the build; that check is now made once the stamp is known to be held, which is what makes it meaningful. No bound is raised and production code is unchanged.Verification
MeshWeaver.Hosting.TestRelease-warnaserror: 0 warnings, 0 errors.--filter AGatedSweepHearsItsOwnCompileTest: 2/2 pass locally.Pairs-with: none — test-only change; no public surface touched.
🤖 Generated with Claude Code