ci: reuse verified build inputs and rebalance measured workloads

[Abiorh001]

Backend CI repeatedly rebuilds unchanged MinIO inputs and downloads the same Python packages. Its checker delivery workload also exceeded the execution deadline. This change reuses verified build inputs and partitions checker delivery and routing preparation across the existing three project lanes.

MinIO caching uses exact recipe/platform keys, with trusted-main publication only. Every consumer still verifies its run-bound artifact and live provider. Pip caches downloads only; each job installs into a fresh environment. Complete duration diagnostics remain in retained logs.

The first allocation put all checker delivery tests in project A. Run 37809916459, tested merge dbc84f10e1ca1926412856a5f34e4bc91ab49c5d, disproved that placement: A timed out at 1,200.966 seconds after 665/745 tests; B/C completed in 769.294/483.442 seconds. Cleanup completed. Delivery and routing preparation now use the existing three-way exact-node hash instead of exclusive placement.

Validation:

• 124 focused catalogue, runner, evidence and aggregation tests passed in 46.13s; 19 workflow and 29 Commitrail structure tests passed.
• Replaying that hosted 8,803-node inventory preserves every node and execution kind exactly once. Only 127 assignments within the 15 intended modules change. This is historical inventory replay, not current-head full-suite proof.
• Restoring the overloaded predecessor allocation fails the updated partition regression; the repaired allocation passes. Ruff, actionlint, Commitrail, Markdown links and stale-wording/review-contract checks pass.
• Independent CI integrity, security, QA and test-delta review covers the frozen repair. Fresh hosted CI remains required; runtime forecasts are not passing evidence.

Review target: base/merge-base 36e8a615f01801cecd6ccf83d7411235ab1cfa8f; head b5401577ba510fe49553b791e0c36ab82dd88c2a; tree 8192d7ff62480ad2551d25af3535aad7dcb31bf3.

All nine jobs, test completeness, deadlines, isolation, UUID collection, failure propagation and aggregate custody remain. No product source, migration or branch-protection changes. Roadmap assessment: this changes CI execution and operator guidance, with no change to product capability or exposure.

Summary by CodeRabbit

• CI Improvements
  • Backend CI now reuses verified MinIO image and Python package caches across eligible runs. Cache updates are saved only from qualifying pushes to the main branch.
  • Test reports now include timing information for all tests, and backend test workloads are distributed across the project lanes.
• Documentation
  • Updated backend testing guidance with lane assignments, rerun instructions, and cache behavior.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

• Configuration used: defaults
• Review profile: CHILL
• Plan: Advanced
• Run ID: a155439b-7ca9-4461-a044-2b6ac1c1e975

📥 Commits

Reviewing files that changed from the base of the PR and between 3b9fbf6 and b77a1e9.

📒 Files selected for processing (10)

• .commitrail/changes/backend-ci-cache-and-timing.md
• .github/workflows/backend.yml
• CONTRIBUTING.md
• backend/scripts/run_test_lanes.py
• backend/scripts/test_lane_catalogue.py
• backend/tests/test_ci_lane_catalogue.py
• backend/tests/test_ci_test_lanes.py
• docker/minio/README.md
• docs/operations_backend_testing.md
• scripts/test_lightweight_agent_gates.py

Included review availability: This review used your included allowance. Your plan provides up to 1 included review per hour; 0 remain after this review.

---

📝 Walkthrough

Walkthrough

Backend CI now uses content-keyed MinIO and pip caches, redistributes checker-delivery and routing-authorization tests across the three project lanes, and reports durations for all tests. Tests and documentation cover the updated cache rules, lane assignments, and rerun guidance.

Changes

Backend CI

Layer / File(s)	Summary
Cache keys and publication .github/workflows/backend.yml, docker/minio/README.md, scripts/test_lightweight_agent_gates.py, .commitrail/changes/backend-ci-cache-and-timing.md	The workflow restores MinIO and pip caches using keys based on runner and source inputs. Cache misses are saved only on pushes to main. The MinIO image is verified before saving, and each job continues to use fresh installs.
Lane allocation and diagnostics backend/scripts/test_lane_catalogue.py, backend/scripts/run_test_lanes.py, backend/tests/test_ci_lane_catalogue.py, backend/tests/test_ci_test_lanes.py, docs/operations_backend_testing.md, CONTRIBUTING.md, .commitrail/changes/backend-ci-cache-and-timing.md	Checker-delivery and routing-authorization preparation tests move into the three project lanes. Pytest reports durations for all tests. Tests and documentation describe lane ownership and rerun guidance.

Priority: ⬇️ Low

Estimated code review effort: 3 (Moderate) | ~25 minutes

Change: Other

Suggested reviewers: durutheguru

Merge Risk: ⚪ Minimal · up to b77a1

No identified issue blocks merging after normal CI checks. Hosted CI should still confirm the expected runtime and cache behavior.

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage		Docstring coverage is 13.33% which is insufficient. The required threshold is 80.00%. Docstring coverage is scoped to functions touched by this diff. Analyzed 15 functions across 5 files. (5 skipped: …	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Title check		The title clearly and concisely describes the two primary changes: reusing verified CI inputs and rebalancing measured workloads.
Description check		The description is detailed and on topic. It explains the goal, cache design, workload rebalance, measured failure, validation results, scope, and remaining need for fresh hosted CI. It does not repro…
Linked Issues check		Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check		Check skipped because no linked issues were found for this pull request.

Full details: Docstring Coverage

Explanation

Docstring coverage is 13.33% which is insufficient. The required threshold is 80.00%. Docstring coverage is scoped to functions touched by this diff. Analyzed 15 functions across 5 files. (5 skipped: 5 unsupported.)

• Fix all pre-merge checks with AI

✨ Finishing Touches 💡 1

📝 Generate docstrings 💡

• Commit to this branch
• Create a new PR

🧪 Generate unit tests (beta)

• Commit to this branch
• Create a new PR

• Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

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