ci: cover protected runner queue time

[chaofengw-nv]

Background

The first post-merge live validation on the dedicated external-contributor PR #1056 exposed a real runner-queue timeout. Public bridge run 33080320518 correctly correlated private run 33080349314 by nonce, but its Wait for the exact Internal CI result step failed after 56m53s while the selected fast_foundation_stereo GPU job was still queued with no steps started. The failure was therefore in the bridge waiting budget, not in nonce correlation, artifact generation, or failure-log rendering.

Exit Criteria

• The public bridge wait budget covers protected runner queueing plus the longest supported Internal model proof.
• The bridge retains time after the private run completes to download, validate, print, upload, and publish the failure result.
• The exact-run nonce correlation and fail-closed behavior remain unchanged.
• A deterministic workflow test locks down the timeout and polling budget.

Implementation

• Raises the trusted dispatch job timeout from 60 to 360 minutes.
• Raises the exact private-run polling window from 220 to 1,400 15-second polls, or 350 minutes.
• Reserves the final 10 minutes for artifact transfer, closed-contract validation, public .log publication, and status update.

Change categories

• Model or runtime behavior
• Public API
• ABI
• Bundle or artifact format
• Dependencies
• Documentation only
• CI or developer tooling

Validation

Commands and Results

• Regression test before the workflow change: failed because the dispatch timeout was 60 rather than 360 minutes.
• Focused bridge test after the change: 1 passed.
• python3 -m tools.community_ci source-quality --base github/main: passed; 158 tests passed, changed-file lint/formatting passed, and the complexity gate passed.
• git diff --check: passed.

Hardware, Environment, and Revisions

• Source head: 870bf4087921acaac9c8f4a8edf138fa6093b8b8.
• Source base: 7aa0b211e52ad17f6d827d4c33022b36130e7f01.
• Live evidence: public bridge 33080320518, private Internal run 33080349314, Source PR test(stereo): cover DHWC8 and probe GPU CI logs #1056 head d3b8f6650f110773a6a54c6d37fcf00203bffb80.
• The private GPU job remained queued for more than 55 minutes with no steps, reproducing the budget failure without executing contributor code in the public workflow.

Not Run / Remaining Gaps

The complete public failure-log handoff could not run because the original bridge exhausted its 55-minute polling window before a protected GPU runner became available. After this change reaches main, rerun only #1056 and verify the nonce-correlated artifact, printed .log, status target, root CTest failure block, and sensitive-data checks end to end.

Notes For Future Readers

The longest Internal model proof currently has a 180-minute job timeout. The 350-minute polling window covers that execution budget plus substantial protected-runner queueing while staying within GitHub's six-hour hosted-job limit. If Internal model timeouts grow beyond 180 minutes, revisit both budgets together.

Risk level

• Low
• Medium
• High

Risk rationale: this changes only how long the trusted bridge waits for an already exact, nonce-correlated private run. Permissions, payload contents, publication checks, and status context are unchanged.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Enterprise

Run ID: b8ea547a-b0e7-4f1e-b961-428f7832d1e7

📥 Commits

Reviewing files that changed from the base of the PR and between 7aa0b21 and 870bf40.

📒 Files selected for processing (2)

• .github/workflows/internal-ci-bridge.yml
• tests/tools/test_github_actions_ci.py

Included review availability: Your plan provides up to 12 included reviews per hour; 11 remain after this review.

---

📝 Walkthrough

Summary by CodeRabbit

• Bug Fixes

  • Increased the CI bridge timeout to support longer-running validation jobs.
  • Extended polling time so builds have more time to complete before being marked as failed.

• Tests

  • Updated automated checks to verify the extended timeout and polling limits.

Walkthrough

The internal CI bridge now supports a 360-minute dispatch timeout and up to 1,400 polling iterations. Tests verify both workflow limits.

Changes

Internal CI bridge timing

Layer / File(s)	Summary
Extend CI timing and validate limits .github/workflows/internal-ci-bridge.yml, tests/tools/test_github_actions_ci.py	The dispatch timeout increased to 360 minutes. The exact Internal CI run polling limit increased to 1,400 iterations. Tests assert both values.

Estimated code review effort: 1 (Trivial) | ~5 minutes

Merge Risk: ⚪ Minimal · up to 870bf

This PR extends the trusted CI bridge wait budget and adds deterministic coverage without changing permissions, payloads, or publication behavior; no actionable merge-blocking risk remains after normal checks and review.

Suggested reviewers: yifeif-nv

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 0.00% which is insufficient. The required threshold is 80.00%. Docstring coverage is scoped to functions touched by this diff. Analyzed 1 functions across 1 files. (1 skipped: 1 …	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Description check	✅ Passed	The description covers the background, exit criteria, implementation, change category, validation results, revisions, remaining gaps, risk, and rollout context. It clearly identifies the end-to-end va…
Title check	✅ Passed	The title is concise and accurately identifies the primary change: extending CI coverage for protected runner queue time.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

Full details: Description check

Explanation

The description covers the background, exit criteria, implementation, change category, validation results, revisions, remaining gaps, risk, and rollout context. It clearly identifies the end-to-end validation that remains after merge.

Full details: Docstring Coverage

Explanation

Docstring coverage is 0.00% which is insufficient. The required threshold is 80.00%. Docstring coverage is scoped to functions touched by this diff. Analyzed 1 functions across 1 files. (1 skipped: 1 unsupported.)

---

_{Comment @coderabbitai help to get the list of available commands.}