docs(rfc): rfc 0034 — d1 re-scope on measured premises

[jensholdgaard]

drafted RFC 0034, rewritten on true premises after the first draft's core claim was refuted by measurement (a sequencing story the RFC records deliberately): D1's per-core metric contradicts its own per-node falsifier, and per-core is the wrong axis twice over — the per-tenant miner is sequential by design (§3.7), and the node ceiling was a global serialization (§9.21), removed by RFC 0035 Design A (132,289 lines/s measured, §9.22).

Recast: must-win = per-node ≥ 100k lines/s on baseline-8vcpu-32gib (multi-tenant, one shared commit stream; Design A clears with ~32% margin; the old per-core number gets a per-node home), p99 ≤ 200 ms at the sustained rate, per-core + per-tenant rates demoted to recorded diagnostics (RFC 0011 pattern), D2 untouched, falsifier retained verbatim. The bar asserts only once RFC 0035 reaches green — the prototype number calibrates, it does not verdict.

Doc-only; mdbook verified.

🤖 Generated with Claude Code

https://claude.ai/code/session_01WQY9wfrfRggqSpMLH8Xj3Y

Summary by CodeRabbit

• Documentation
  • Added RFC 0034, redefining the D1 performance gate around sustained per-node ingest throughput under multi-tenant load.
  • Clarified the p99 ingest acknowledgment target and introduced diagnostic reporting for per-core and per-tenant rates.
  • Documented updated acceptance scenarios, testing guidance, open questions, and related references.
  • Added RFC 0034 to the documentation index.

[coderabbitai]

Warning

Review limit reached

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

Next review available in: 52 minutes

Enable usage-based reviews in Billing to review now. Otherwise, wait until the next included review is available.
You're only billed for reviews past your plan's rate limits ($0.25/file).

How can I continue?

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.

To avoid repeated limits, reduce automatic review volume by pausing incremental auto-reviews earlier, using label-based review opt-in, excluding WIP or generated PR titles, or requesting reviews manually when the PR is ready. If your team needs uninterrupted high-volume reviews, an organization admin can enable usage-based reviews.

How do review limits work?

CodeRabbit enforces per-developer PR review limits for each organization. Most developers receive the normal plan review availability.

For paid Pro and Pro+ PR reviews, CodeRabbit uses adaptive limits for sustained high-volume activity. When a developer's recent PR review activity reaches the 95th percentile or higher among CodeRabbit users, additional reviews become available more gradually as earlier reviews age out of the rolling window.

Please refer docs for additional details.

Review details

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro Plus

Run ID: 25cf34bf-4741-46c2-b715-1e616bd0a35d

📥 Commits

Reviewing files that changed from the base of the PR and between bbe5206 and f34bdd1.

📒 Files selected for processing (1)

• docs/rfcs/0034-d1-rescope.md

📝 Walkthrough

Walkthrough

Adds RFC 0034 and indexes it in the documentation. The RFC changes D1 from a per-core to a per-node sustained throughput gate, preserves the ingest-ack SLO, defines diagnostics and acceptance scenarios, and documents testing, open questions, and references.

Changes

D1 re-scope documentation

Layer / File(s)	Summary
Revised D1 gate definition docs/rfcs/0034-d1-rescope.md	Defines D1 as sustained per-node throughput of at least 100k lines/s on the baseline configuration, retains the p99 ingest-ack limit, and makes per-core and per-tenant rates diagnostic only.
Acceptance and testing criteria docs/rfcs/0034-d1-rescope.md	Documents rejected alternatives, updated acceptance scenarios, RFC 0035 gating, and the sustained soak-run validation approach.
References and documentation navigation docs/rfcs/0034-d1-rescope.md, docs/SUMMARY.md	Adds RFC metadata, open questions, related references, and the RFC index entry.

Estimated code review effort: 2 (Simple) | ~10 minutes

Possibly related PRs

• jensholdgaard/ourios#568: Documents D1 capacity and throughput evidence on the same baseline configuration.
• jensholdgaard/ourios#572: Introduces the RFC 0035 ingest concurrency and WAL changes referenced by RFC 0034.

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Description check	⚠️ Warning	The description explains the change, but it doesn't follow the required template headings or include a Related section or checklist items.	Reformat the PR description with Summary, Related, and Checklist sections, and include the RFC link plus all checklist items.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title clearly identifies the RFC 0034 D1 re-scope change and matches the main theme of the PR.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
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.

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch rfc-0034-d1-rescope-v2

---

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.}

[Copilot]

Pull request overview

Adds a new tuning RFC (RFC 0034) to document a measured re-scope of the D1 ingest throughput gate from a per-core metric to a per-node capacity target on baseline hardware, and wires it into the mdBook navigation. This fits the codebase’s RFC-driven workflow by updating the benchmarks/gates framing without changing implementation.

Changes:

• Added docs/rfcs/0034-d1-rescope.md (status drafted) describing the D1 re-scope rationale, design, and acceptance criteria.
• Linked RFC 0034 in docs/SUMMARY.md so it appears in the rendered mdBook.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated no comments.

File	Description
docs/SUMMARY.md	Adds RFC 0034 to the mdBook table of contents.
docs/rfcs/0034-d1-rescope.md	Introduces RFC 0034 documenting the D1 gate recalibration and its acceptance criteria.

---

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

[coderabbitai]

Actionable comments posted: 2

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Inline comments:
In `@docs/rfcs/0034-d1-rescope.md`:
- Around line 166-167: Update the asserting-run tenant-count item in RFC 0034 so
it no longer remains an unresolved choice between N = cores and a tenant-curve
knee. Define a fixed deterministic N, or specify a reproducible knee-selection
procedure using the RFC 0035 green run, and explicitly enforce N ≥ cores.
- Around line 84-93: Update the D1 acceptance criterion around the “below
saturation” clause to define an observable proof, using either a measurable
throughput/headroom threshold or an explicit load-sweep rule. Ensure the
asserting run records the required offered-versus-achieved throughput evidence
and applies the same rule when evaluating the p99 ingest-ack limit.

🪄 Autofix (Beta)

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro Plus

Run ID: 052e4be9-cd80-466f-b6b1-cbdc57727361

📥 Commits

Reviewing files that changed from the base of the PR and between 5d1fbe2 and bbe5206.

📒 Files selected for processing (2)

• docs/SUMMARY.md
• docs/rfcs/0034-d1-rescope.md

[Copilot]

Pull request overview

Copilot reviewed 2 out of 2 changed files in this pull request and generated 1 comment.

[Copilot]

Pull request overview

Copilot reviewed 2 out of 2 changed files in this pull request and generated no new comments.

Comments suppressed due to low confidence (1)

docs/rfcs/0034-d1-rescope.md:4

• RFC includes §5 Acceptance criteria and §6 Testing strategy, which per docs/rfcs/README.md lifecycle rules corresponds to status: specified (Drafted RFCs should only fill §§1–4 and §§7–8). Keeping status: drafted here makes the RFC lifecycle inconsistent for reviewers/tooling.

status: drafted