fix(teardown): a hub disposing with its OWN deferred delivery strands nobody — Debug, not Error (#4178) by rbuergi · Pull Request #4191 · Systemorph/MeshWeaver

rbuergi · 2026-09-13T14:53:48Z

The defect

MessageService.Dispose logs event 7301 at Error for every deferred delivery it discards, saying "the message is NOT processed; the sender is answered ShuttingDown".

For a self-addressed delivery that sentence is false. One method down, NackThroughParent declines exactly this delivery with NACK_DECLINED reason=sender-is-self — there is nowhere to post the answer, because the pending-response registry that would resolve it belongs to this hub and is cancelled in the same Dispose. The reader is told by a different and better route (CancelCallbacks errors those subjects with "Hub … was disposed before the response arrived"). So the Error alleged stranded work, named an answer that was never sent, and told whoever read it to go find a producer that had done nothing wrong.

The measurement

76 Error lines per plugin-gate shard (Plugins run 34731952463, Gate shard 1/1, same line on every shard measured 2026-09-13), every one a $model-probe/{guid} hub discarding its own GetDataRequest parked behind [DataContextInit,MeshNodeInit]:

[DISPOSE-DISCARD] Hub $model-probe/f27ab9f9… is disposing with GetDataRequest
(id=beQyWMaX60eN-jRJQeKllw, from $model-probe/f27ab9f9…) still deferred; initialization gates
closed at deferral: [DataContextInit,MeshNodeInit] — the message is NOT processed; the sender is
answered ShuttingDown.

Sender == target == the hub. A transient node probe is created, read once and disposed — TransientNodeProbe states that outright, and MessageHub.HandleInitialize already classifies the same probe's dispose-during-init as a recognized shutdown rather than a failure (#1122–#1125). Its own-node read collapsing onto its synthetic address is the shape TransientProbeAddresses exists for; two other read seams (GetMeshNodeOutcome, MeshNodeStreamCache.GetStreamRaw) already answer it directly. This is the third, reached as a raw deferral.

The change

One predicate at the discard site. When delivery.Sender is this hub's own address, the line is logged at Debug with a sentence that no longer claims an answer went out; it still names the message type, id, the gates it sat behind and the run level. A delivery from any other sender strands a real waiter on a transient NACK and stays an Error, unchanged.

The rule follows the fact, not the probe — it is keyed on "is anyone outside waiting", the same thing NackThroughParent already decides, not on a probe-address special case.

Why this is not a silenced fault

Nothing is suppressed and no bound was widened: the delivery is answered exactly as before (it was never answered for this shape), the site still reports, and the external-sender Error — the one with a victim who can act — is untouched. What changed is a verdict that was measurably false for one shape.

The alternative considered and not taken: guard the third read seam so a probe's own-address GetDataRequest is answered immediately instead of parked. That is a larger change to what probes read, with no way to know which probes depend on the parked read completing — and it would still leave the Error wrong for every other self-addressed discard.

Negative controls — both watched failing

control	what breaks	assertion that reds
predicate → always false (pre-fix: everything Error)	the self-addressed case stops being reported at Debug	"the discard must STILL be reported and still name the gate that held it — a reclassification that stops reporting would be a silenced fault, not a classification"
predicate → always true (everything Debug)	the external-sender Error disappears	"a delivery from ANOTHER hub strands a real waiter on a transient NACK — that stays an Error"

The first control is the one that matters: it proves the new test cannot pass by the line vanishing. The log capture was widened to record the level, so "Error is gone" and "the line is gone" are two different readings.

Verification

dotnet build -c Release -warnaserror (whole solution): 0 Warning(s), 0 Error(s)
MeshWeaver.Messaging.Hub.Test: 436 passed, 0 failed
MeshWeaver.Documentation.Test: 447 passed, 0 failed (includes DocumentationLinkIntegrityTest for the two doc edits)
TestTimeoutLiteralRatchetGuard reported the new test's literal Timeout = 30_000 and was obeyed by converting the site, not raising the baseline — the test's waits already carry TestTimeouts.Convergence.

Docs

TeardownLayers.md — "What a hub still discards, and why that is an Error too" gains the self-addressed split, with the sender-is-self decline as the reason.
What's New: A hub disposing with its own question unanswered is no longer reported as lost work (Category: Fix).

Cross-repo

No public type or member removed, no interface member added, no i18n catalog key touched — Pairs-with:, Implementers: and Mirror-sync: do not apply.

Fixes #4178

🤖 Generated with Claude Code

… nobody — Debug, not Error (#4178) `MessageService.Dispose` logged event 7301 at Error for every deferred delivery it discarded, with the sentence "the sender is answered ShuttingDown". For a SELF-ADDRESSED delivery that sentence is false: `NackThroughParent` declines it one method down with `NACK_DECLINED reason=sender-is-self`, because the pending-response registry that would resolve it is this hub's own and is cancelled in the same disposal. The reader is told by `CancelCallbacks` instead ("Hub … was disposed before the response arrived") — a better report, to the same reader. So the Error alleged stranded work, named an answer that was never sent, and pointed the reader at a producer that had done nothing wrong. Measured (#4178): 76 Error lines per plugin-gate shard, every one a `$model-probe/{guid}` hub discarding its own `GetDataRequest` parked behind [DataContextInit,MeshNodeInit] — the lifetime `TransientNodeProbe` declares by design ("created, read once, disposed"), and which `MessageHub.HandleInitialize` already classifies as a recognized shutdown rather than a failure (#1122-#1125). The rule is the FACT, not the probe: when the sender IS this hub, nobody outside is waiting, so the discard is teardown-internal. The line still fires and still names the message, id, gates and run level — at Debug, with a sentence that no longer claims an answer went out. A delivery from any OTHER sender still strands a real waiter on a transient NACK and stays an Error, unchanged. Both halves pinned, each watched failing: - widen the predicate to always-true -> the external-sender test reds for LOSING the Error - narrow it to always-false -> the self-addressed test reds for STOPPING to report The second control is the one that matters: a reclassification that quietly stopped reporting would be a silenced fault, not a classification. Docs: TeardownLayers.md gains the split; What's New entry (Fix). Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

Copilot

🟡 Changes recommended

The diagnostic wording is too specific, and the regression test does not cover pending-response teardown or verify the corrected message.

Get a fresh assessment by requesting another Copilot review.

Pull request overview

This PR adjusts teardown logging for self-addressed deferred messages, keeping externally originated deliveries at Error level.

Changes:

Reclassifies self-addressed discards as Debug.
Adds regression coverage for log levels and messages.
Updates teardown documentation and release notes.

File summaries

File	Summary
`test/MeshWeaver.Messaging.Hub.Test/DeferredDeliveryNackedOnDisposeTest.cs`	Adds self-addressed teardown logging coverage.
`src/MeshWeaver.Messaging.Hub/MessageService.cs`	Updates deferred-discard classification and diagnostic text.
`src/MeshWeaver.Documentation/Data/WhatsNew/2026-09-13-a-hub-disposing-with-its-own-question-unanswered-is-not-an-error.md`	Adds the release note.
`src/MeshWeaver.Documentation/Data/Architecture/TeardownLayers.md`	Documents the revised teardown behavior.

Review details

Suppressed comments (1)

test/MeshWeaver.Messaging.Hub.Test/DeferredDeliveryNackedOnDisposeTest.cs:165

This does not reproduce the production self-addressed question: Post is fire-and-forget, so no response subject is registered and NackThroughParent exits with NACK_DECLINED reason=not-awaited, never sender-is-self; CancelCallbacks also has nothing to cancel. Use Observe with an error subscriber to register the callback before posting, otherwise this test can pass without exercising the pending-response teardown path described by the PR.

        selfTalker!.Post(new GatedRequest(), o => o.WithTarget(SelfTalkerAddress));

Files reviewed: 4/4 changed files
Comments generated: 2
Review effort level: Lite

💡 Configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

+                    + "disposal. Teardown-normal (a transient node probe is created, read once and "
+                    + "disposed by design); not a discard of anybody else's work.",


+        log.At(LogLevel.Debug).Should().Contain(m => m.Contains("self-gate-never-opens"),
+            "the discard must STILL be reported and still name the gate that held it — a "
+            + "reclassification that stops reporting would be a silenced fault, not a classification");


github-actions · 2026-09-13T14:59:00Z

Test Results (shard 1)

512 tests 512 ✅ 39s ⏱️
1 suites 0 💤
1 files 0 ❌

Results for commit 3c42160.

♻️ This comment has been updated with latest results.

github-actions · 2026-09-13T14:59:02Z

Test Results (shard 0)

262 tests 262 ✅ 2m 8s ⏱️
1 suites 0 💤
1 files 0 ❌

Results for commit 3c42160.

github-actions · 2026-09-13T14:59:56Z

Test Results (shard 3)

1 263 tests 1 263 ✅ 1m 42s ⏱️
2 suites 0 💤
2 files 0 ❌

Results for commit 3c42160.

github-actions · 2026-09-13T15:02:29Z

Test Results (shard 5)

4 files 4 suites 3m 56s ⏱️
2 082 tests 2 082 ✅ 0 💤 0 ❌
2 086 runs 2 086 ✅ 0 💤 0 ❌

Results for commit 3c42160.

github-actions · 2026-09-13T15:02:57Z

Test Results (shard 4)

1 576 tests 1 576 ✅ 4m 39s ⏱️
3 suites 0 💤
3 files 0 ❌

Results for commit 3c42160.

github-actions · 2026-09-13T15:06:00Z

Test Results (shard 2)

1 530 tests 1 530 ✅ 7m 45s ⏱️
3 suites 0 💤
3 files 0 ❌

Results for commit 3c42160.

github-actions · 2026-09-13T15:06:23Z

Test Results

14 files 14 suites 20m 51s ⏱️
7 225 tests 7 225 ✅ 0 💤 0 ❌
7 229 runs 7 229 ✅ 0 💤 0 ❌

Results for commit 3c42160.

♻️ This comment has been updated with latest results.

Copilot AI lite review requested due to automatic review settings September 13, 2026 14:53

Copilot started reviewing on behalf of rbuergi September 13, 2026 14:54

View session

rbuergi mentioned this pull request Sep 13, 2026

[DISPOSE-DISCARD] 76 Error-level lines per plugin-gate shard: $model-probe hubs dispose with a GetDataRequest still deferred behind [DataContextInit,MeshNodeInit] #4178

Closed

Copilot AI reviewed Sep 13, 2026

View reviewed changes

rbuergi marked this pull request as ready for review September 13, 2026 15:11

meshweaver-cloud Bot enabled auto-merge September 13, 2026 15:11

meshweaver-cloud Bot added this pull request to the merge queue Sep 13, 2026

Merged via the queue into main with commit 7a3f171 Sep 13, 2026
57 of 59 checks passed

rbuergi deleted the fix/4178-self-addressed-dispose-discard branch October 10, 2026 13:10

Navigation Menu

Repository navigation

fix(teardown): a hub disposing with its OWN deferred delivery strands nobody — Debug, not Error (#4178) - #4191

fix(teardown): a hub disposing with its OWN deferred delivery strands nobody — Debug, not Error (#4178)#4191
meshweaver-cloud[bot] merged 1 commit into
mainfrom
fix/4178-self-addressed-dispose-discard

rbuergi commented Sep 13, 2026

Uh oh!

Copilot AI left a comment

Uh oh!

github-actions Bot commented Sep 13, 2026 •

edited

Loading

Uh oh!

github-actions Bot commented Sep 13, 2026

Uh oh!

github-actions Bot commented Sep 13, 2026

Uh oh!

github-actions Bot commented Sep 13, 2026

Uh oh!

github-actions Bot commented Sep 13, 2026

Uh oh!

github-actions Bot commented Sep 13, 2026

Uh oh!

github-actions Bot commented Sep 13, 2026 •

edited

Loading

Uh oh!

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

		+ "disposal. Teardown-normal (a transient node probe is created, read once and "
		+ "disposed by design); not a discard of anybody else's work.",

Conversation

rbuergi commented Sep 13, 2026

The defect

The measurement

The change

Why this is not a silenced fault

Negative controls — both watched failing

Verification

Docs

Cross-repo

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

🟡 Changes recommended

Suppressed comments (1)

Uh oh!

github-actions Bot commented Sep 13, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Test Results (shard 1)

Uh oh!

github-actions Bot commented Sep 13, 2026

Test Results (shard 0)

Uh oh!

github-actions Bot commented Sep 13, 2026

Test Results (shard 3)

Uh oh!

github-actions Bot commented Sep 13, 2026

Test Results (shard 5)

Uh oh!

github-actions Bot commented Sep 13, 2026

Test Results (shard 4)

Uh oh!

github-actions Bot commented Sep 13, 2026

Test Results (shard 2)

Uh oh!

github-actions Bot commented Sep 13, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Test Results

Uh oh!

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

github-actions Bot commented Sep 13, 2026 •

edited

Loading

github-actions Bot commented Sep 13, 2026 •

edited

Loading