Skip to content

Record bounded Cell acceptance progress - #1009

Open
zlareb1 wants to merge 16 commits into
mainfrom
fix/cell-provenance-phase-evidence
Open

zlareb1 wants to merge 16 commits into
mainfrom
fix/cell-provenance-phase-evidence

Conversation

@zlareb1

@zlareb1 zlareb1 commented Nov 20, 2026 •

Copy link
Copy Markdown
Member

Related: #1041
Related: #1072
Related: #1077
Related: #1087
Related: #1091

Changelog category

  • CI Fix or Improvement

Changelog entry

Not required.

Why

An interrupted Cell provenance proof needs its last observed page and validation phase. Bounded numeric checkpoints retain that evidence beside the acceptance result. Selected HTTP helper timings show request work at the existing sample boundaries.

Scope and risk

The harness exports bounded provenance progress, observed remaining work budget and request-phase timings. Only the existing six sampled requests receive the nine HTTP boundary observations. Wrappers preserve arguments, return values, cancellation and error identity, and restore helpers after startup failure or completion. Records contain numeric checkpoints and hashed query tags.

The proof-default comparison uses the related canonical-mapping correction to bound encoder setup while retaining every field and JSON type check.

Qualification controls exercise the isolated replicated-erasure lease table already on main, decision-tag filtering for the SCIM query recorder and a private cleanup clock. Missing-lease refusal, unbounded-query negative controls, original cleanup deadlines and original assertions are retained. The existing lease-table fixture remains unchanged.

The graph-stats cancellation fixture includes the related private idle-clock correction. Its original pass deadline, signal watchdog, post-cancellation waits, statement-count freeze and test bodies remain intact. Pure graph calculations use their original functions through the fixture's inline thread seam; this fixture does not qualify real executor or wall-clock latency.

Workload, pages, resource census, global deadlines, latency and throughput bounds are retained. Latency, throughput and complete acceptance require hosted validation.

Compatibility, rollout and rollback

  • Wire, production server behavior, schema, settings and operator steps: none.
  • The diagnostic harness writes bounded progress artifacts beside its existing acceptance output.
  • Rollback: revert the harness and fixture changes. No persisted server data changes.

Prior-source qualification

The prior feature command records name their recorded heads but lack per-run immutable working-source hashes. Their outcomes do not independently establish the executed feature bytes. Committed-file comparisons establish source equality only.

Historical command records report that the graph-budget, provenance-progress and selected-request-phase controls passed. Original progress fields, failure identity, workload, deadline and privacy bounds remain unchanged. The earlier integration retrieval failure remains real with an unproven cause. Historical Cell success on an earlier head does not qualify this head or establish a latency repair.

Current main adds stored review provenance to the proof protocol. This composition retains its 62 memory fields, 44 sparse defaults and stored-provenance requirement, and extends the default-field refusal coverage to the new field. Other original source entries retain their exact modes and bytes. Current main supplies the person-query observations and remaining merged prerequisites. Earlier records include scoped Ruff, formatting, public prose and committed-range whitespace results. Their commands and scoped outcomes remain historical. No previous review or CI result qualifies this head; a fresh independent entire review and all required full exact-head hosted workflows must pass before merge. No runtime, provider call, performance probe or hosted retry was run.

Tests

This composition includes main's grep response fixture and private retirement clock controls. The committed tree preserves every prior concern entry outside those three main paths, with its exact blob and mode. The three main entries match main exactly.

Current local checks pass for the retirement clock's blocked-wait control, syntax of both grep test files, scoped Ruff and formatting of the retirement unit file, public prose and whitespace in the three-path composition. Feature qualification listed above remains historical. Committed-source equality is separately verified; those checks were not repeated unless named here. No prior whole review or hosted CI verdict qualifies this head. Full new exact-head hosted CI and a fresh independent whole review remain required before merge. Existing failures and unproven causes remain unresolved; no performance or runtime repair is claimed.

AI assistance in the code

  • AI-generated: Codex wrote the changes and controls, then checked bounded encoding and diagnostics, refusal paths, retained assertions and task cleanup.

  • I can explain every line of this diff and how it interacts with the rest of Loom.

Checklist

  • Acceptance thresholds, workload and deadlines are retained.
  • No production source, generated interfaces, migrations or operator settings change.
  • Commits are signed off.
  • Full exact-head GitHub CI passes.

@zlareb1 zlareb1 changed the title Record bounded Cell provenance progress Record bounded Cell acceptance progress Nov 20, 2026
@zlareb1
zlareb1 force-pushed the fix/cell-provenance-phase-evidence branch from 247b780 to 2de349e Compare November 20, 2026 23:06
@vercel

vercel Bot commented Nov 20, 2026 •

Copy link
Copy Markdown

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

Project Deployment Actions Updated
loom-docs Ready Ready Preview Nov 21, 2026 11:33am UTC

Request Review

Exercise each default-field refusal and JSON numeric distinction before consolidating repeated default comparisons.

Signed-off-by: zlareb1 <lareb.zafar@clickhouse.com>
(cherry picked from commit 73ef63b)
Signed-off-by: zlareb1 <lareb.zafar@clickhouse.com>
Keep every default-field and JSON type check while avoiding a separate encoder setup for each field.

Signed-off-by: zlareb1 <lareb.zafar@clickhouse.com>
(cherry picked from commit 0bfc712)
Signed-off-by: zlareb1 <lareb.zafar@clickhouse.com>
Shared client schema reads run with an independent context. Record only decision-tagged statements while retaining unbounded decision reads as failures.

Signed-off-by: zlareb1 <lareb.zafar@clickhouse.com>
(cherry picked from commit 57327d5)
Signed-off-by: zlareb1 <lareb.zafar@clickhouse.com>
Keep the blocked close, deadline observers and loop timers on one private
clock. A controlled independent clock advance reproduces the elapsed-time
dependency while the original cleanup and refusal assertions stay intact.

Signed-off-by: zlareb1 <lareb.zafar@clickhouse.com>
(cherry picked from commit 4b03db4)
Signed-off-by: zlareb1 <lareb.zafar@clickhouse.com>
@zlareb1
zlareb1 force-pushed the fix/cell-provenance-phase-evidence branch from 2de349e to 9765f69 Compare November 21, 2026 00:08
zlareb1 added a commit that referenced this pull request Nov 21, 2026
Related: #1009

### Changelog category (leave one):

- CI Fix or Improvement (changelog entry is not required)

### Changelog entry

The graph-stats cancellation fixture uses an idle clock for its pass deadline.

### Why

The pass timeout starts before the cheap legs reach the blocked unit-risk query. Unrelated scheduling time can exhaust that allowance before the fixture observes its in-flight signal. The hosted failure occurs at the two-second signal watchdog; the actual scheduling cause of that run is unknown. A controlled monotonic jump independently reproduces this wall-clock dependency.

### Scope and risk

One existing test module changes. The cancellation case and its controls use a private function loop with the existing manual clock. Pure calculations run inline through the fixture's thread-offload seam, with the original functions, arguments and results. The fixture therefore controls both loop time and executor scheduling while the real graph-stats pass, repository fake and cancellation paths run.

The original test body retains its 0.2-second pass deadline, two-second signal watchdog, five 0.01-second post-cancellation waits, two namespaces, statement-count freeze and completed cheap-leg assertions. Teardown cancels and drains unexpected tasks and rejects leaks. A negative control observes cancellation of a genuinely blocked edge query at the unchanged 0.2-second deadline and checks the complete 0.25-second logical timeline through the original freeze assertions. Independent review must assess the fixture's scheduling coverage.

### Compatibility, rollout and rollback

- Wire: no operations, fields or defaults change.
- Operator steps: none. No production code, configuration, migration or stored data changes.
- Rollback: reverting restores the prior test clock.

### Tests

Prior source qualification on Python 3.12 with the declared development and code extras showed that the controlled scheduler regression fails its 0.25-second jump case and passes its zero-jump control before the fixture fix. The identical command passed both cases afterward:

```sh
LOOM_ENV_FILE= .venv/bin/python -m pytest tests/unit/test_code_graph_stats_pass_budget.py -k pass_deadline_fixture_ignores_external_scheduler_elapsed -q
```

On the composed head, the complete directly related module passes all 24 cases, including the blocked-query deadline control. Scoped Ruff, Python formatting and whitespace checks pass. Full exact-head hosted CI and fresh independent whole review remain required before merge. No database, provider, runtime or benchmark was started.

### AI assistance in the code (leave one; writing this description with AI is fine):

- AI-generated: Codex diagnosed the timing dependency, wrote the scoped fixture and checked preservation of the original test bodies.

- [x] I can explain every line of this diff and how it interacts with the rest of Loom.

### Checklist

- [ ] Full required hosted CI passes on this head.
- [x] Original deadlines, inputs and behavior assertions are preserved.
- [x] No generated, wire, schema, plugin, frontend or operator documentation change is required.
- [x] Commits are signed off.

Signed-off-by: zlareb1 <lareb.zafar@clickhouse.com>
An interrupted proof needs its last observed page and validation phase. Numeric checkpoints retain that evidence while preserving proof checks, deadlines and the original failure.

Signed-off-by: zlareb1 <lareb.zafar@clickhouse.com>
(cherry picked from commit f43e6d1)
Signed-off-by: zlareb1 <lareb.zafar@clickhouse.com>
Bound HTTP helper timings to the existing six request samples so acceptance failures retain evidence for work outside named spans. Preserve delegation, failure identity and the original workload and limits.

Signed-off-by: zlareb1 <lareb.zafar@clickhouse.com>
(cherry picked from commit fc936c8)
Signed-off-by: zlareb1 <lareb.zafar@clickhouse.com>
Exercise each default-field refusal and JSON numeric distinction before consolidating repeated default comparisons.

Signed-off-by: zlareb1 <lareb.zafar@clickhouse.com>
(cherry picked from commit 73ef63b)
Signed-off-by: zlareb1 <lareb.zafar@clickhouse.com>
Keep every default-field and JSON type check while avoiding a separate encoder setup for each field.

Signed-off-by: zlareb1 <lareb.zafar@clickhouse.com>
(cherry picked from commit 0bfc712)
Signed-off-by: zlareb1 <lareb.zafar@clickhouse.com>
Shared client schema reads run with an independent context. Record only decision-tagged statements while retaining unbounded decision reads as failures.

Signed-off-by: zlareb1 <lareb.zafar@clickhouse.com>
(cherry picked from commit 57327d5)
Signed-off-by: zlareb1 <lareb.zafar@clickhouse.com>
Keep the blocked close, deadline observers and loop timers on one private
clock. A controlled independent clock advance reproduces the elapsed-time
dependency while the original cleanup and refusal assertions stay intact.

Signed-off-by: zlareb1 <lareb.zafar@clickhouse.com>
(cherry picked from commit 4b03db4)
Signed-off-by: zlareb1 <lareb.zafar@clickhouse.com>
@zlareb1
zlareb1 force-pushed the fix/cell-provenance-phase-evidence branch from feb5f08 to be3a061 Compare November 21, 2026 02:44
Signed-off-by: zlareb1 <lareb.zafar@clickhouse.com>
Signed-off-by: zlareb1 <lareb.zafar@clickhouse.com>
Signed-off-by: zlareb1 <lareb.zafar@clickhouse.com>
Signed-off-by: zlareb1 <lareb.zafar@clickhouse.com>
Signed-off-by: zlareb1 <lareb.zafar@clickhouse.com>
Signed-off-by: zlareb1 <lareb.zafar@clickhouse.com>

This branch has not been deployed

No deployments
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant