docs(benchmarks): authoritative baseline thesis-gate results; rfc 0007 validated

[jensholdgaard]

Records the authoritative thesis-gate results from the dedicated baseline-8vcpu-32gib host (8 dedicated vCPU, 32 GiB, local NVMe-class SSD; runs 2026-06-11/12, no CI run id) as docs/benchmarks.md §9.4, and flips RFC 0007 to validated per the maintainer's 2026-06-12 authorization.

Headline results (§9.4)

Gate	Result	Verdict
B1 ≥ 10× vs zstdcat | grep at 1 GiB	v5 34.2× (5.86 ms vs 200.27 ms, 3/6 RGs, exact 11-row agreement); v6 25.4× (8.03 ms vs 203.87 ms, 5/6 RGs, exact 28-row agreement)	PASS (authoritative)
B2 latency ∝ result, not corpus	windowed-1h flat at 1 RG / 4.19–5.92 ms from 735 k → 11.2 M rows, incl. the first HDFS_v1 reading (second corpus family: 28,207 rows, 1/14 RGs, 13 pruned, 5.92 ms); full-span grows 6.84 → 30.19 ms	PASS (authoritative)
C1 100% bit-identical reconstruction	1.000000 on every corpus (v5 1,213,004/1,213,004; v6 1,208,323/1,208,323)	PASS (formal ≥ 1 M-line gate)
C2 convergence ratio ≥ 0.5 at 1 M lines	v5 0.756 / v6 0.760 (v1–v4 abstain, < 1 M lines)	PASS
A1 compression ≥ 3.0× zstd-19	0.439 → 0.751 monotonic (v1 → v5), crossover unobserved	FAIL — the open gate

Why RFC 0007 flips to validated

The docs/verification.md §3 ladder: "Validated. Every thesis-gate in benchmarks.md §7 that the RFC's pillars touch passes on representative corpora." RFC 0007's pillar is the query engine (pillar #3 — DataFusion); the gates it touches are B1 and B2, both now passing on the §1 hardware baseline over ~1 GB+ corpora including a second corpus family (LogHub HDFS_v1, 11.2 M rows). A1's FAIL does not gate this RFC — A1 belongs to the template-mining/compression pillar, measured under RFC 0006, where its §7 escalation is handled. The prior validated-pending checklist resolves: baseline rerun ✓ (§9.4); second corpus family ✓ (HDFS_v1); denser error band — still open, recorded as a non-blocking quality improvement (B1 bands remain 11 / 28 rows).

A1 hardware-sensitivity flag (recorded honestly in §9.4)

The authoritative A1 deltas sit below the ci-runner series at every size (e.g. v5: ourios 23.8× on baseline vs 26.3× on CI, while zstd-19 stayed ~31.7× on both): the ourios writer's output is environment-sensitive — suspected row-group sizing / threading effects on the resulting encodings. Flagged as an open A1 investigation item alongside the structural gap.

Scope & invariants

Doc-only (docs/benchmarks.md, docs/rfcs/0006-bench-harness.md cross-refs, docs/rfcs/0007-querier.md status). No hot path touched; benchmark numbers are recorded, not produced, here. The hardware is identified only by the baseline-8vcpu-32gib tag, per the §1 methodology note added in this PR. Every number was verified against the raw run logs. mdbook build clean; new tables verified in the rendered HTML.

🤖 Generated with Claude Code

[coderabbitai]

Warning

Review limit reached

@jensholdgaard, we couldn't start this review because you've reached your PR review rate limit.

More reviews will be available in 30 minutes. Learn how PR review limits work.

Your organization has run out of usage credits. Purchase more credits in the billing tab to continue.

⌛ How to resolve this issue?

After more reviews become available, a review can be triggered using the @coderabbitai review command as a PR comment. Alternatively, push new commits to this PR.

We recommend that you space out your commits to avoid hitting the rate limit.

🚦 How do rate limits work?

CodeRabbit enforces hourly rate limits for each developer per organization.

Our paid plans include higher PR review limits than trial, open-source, and free plans. In all cases, reviews become available again over time. During sustained high-volume PR review activity, CodeRabbit may temporarily slow when the next review becomes available.

Please see our Fair Usage Limits Policy for further information.

ℹ️ Review info

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro Plus

Run ID: d65b5453-951d-46c4-b81d-c402011c9f81

📥 Commits

Reviewing files that changed from the base of the PR and between 656986a and 3b3b111.

📒 Files selected for processing (3)

• docs/benchmarks.md
• docs/rfcs/0006-bench-harness.md
• docs/rfcs/0007-querier.md

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch docs/baseline-thesis-gates-rfc0007-validated

---

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 and usage tips.}

[jensholdgaard]

@coderabbitai review

[coderabbitai]

✅ Action performed

Review finished.

Note: CodeRabbit is an incremental review system and does not re-review already reviewed commits. This command is applicable only when automatic reviews are paused.

[Copilot]

Pull request overview

Updates the documentation to record the newly-authorized “authoritative baseline” thesis-gate measurements and to advance RFC 0007 (querier/DataFusion) from green to validated in line with the verification ladder.

Changes:

• Add §9.4 “authoritative baseline” results to docs/benchmarks.md and update earlier B1/B2 narrative to reference the baseline run.
• Flip docs/rfcs/0007-querier.md status to validated and replace the prior status-note history with a condensed validation justification.
• Update docs/rfcs/0006-bench-harness.md to reflect that B1/B2 are no longer “future” and are now measured/owned under RFC 0007.

Reviewed changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated 3 comments.

File	Description
docs/rfcs/0007-querier.md	Advances RFC 0007 to validated and documents why B1/B2 satisfy the validation ladder.
docs/rfcs/0006-bench-harness.md	Updates RFC 0006 text to reflect that B1/B2 are now implemented/measured via RFC 0007.
docs/benchmarks.md	Adds authoritative baseline results (§9.4) and updates B1/B2 sections to point at the authoritative measurements.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.