Benchmarks Website Version 3

[connortsui20]

Summary

Rewrites the benchmarks website. Replaces the static data.json.gz model with a single Rust server binary that owns a DuckDB database and accepts POST /api/ingest from CI.

Design

• Single binary: axum + maud (SSR HTML) + DuckDB + Chart.js. All static assets include_bytes!'d.
• 5 fact tables (compression time, query measurement, vector search, RAG, random access). Backup is a file copy.
• Ingest: versioned JSON envelopes, bearer-token gated.
• Migrator ports v2 history forward via a classifier that routes each record to a fact table or skips it with a typed reason.
• Charts/groups slug-addressed, URL round-trip with no DB lookup.
• Routes: /, /chart/{slug}, /group/{slug}, GET /api/chart/{slug}.
• Deploy: one binary, one DuckDB file, one INGEST_BEARER_TOKEN.

UI/UX is still TBD — the relational backend opens up options we didn't have before.

[codspeed-hq]

Merging this PR will degrade performance by 26.23%

⚠️ Unknown Walltime execution environment detected

Using the Walltime instrument on standard Hosted Runners will lead to inconsistent data.

For the most accurate results, we recommend using CodSpeed Macro Runners: bare-metal machines fine-tuned for performance measurement consistency.

⚡ 5 improved benchmarks
❌ 3 regressed benchmarks
✅ 1161 untouched benchmarks
⏩ 138 skipped benchmarks¹

⚠️ Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
⚡	Simulation	patched_take_10k_first_chunk_only	302.4 µs	272 µs	+11.2%
⚡	Simulation	take_10k_dispersed	284.8 µs	239.7 µs	+18.8%
⚡	Simulation	patched_take_10k_adversarial	259 µs	228.6 µs	+13.32%
⚡	Simulation	patched_take_10k_dispersed	316 µs	285.5 µs	+10.69%
⚡	Simulation	take_10k_first_chunk_only	270.8 µs	225.8 µs	+19.92%
❌	Simulation	bitwise_not_vortex_buffer_mut[1024]	307.8 ns	366.1 ns	-15.93%
❌	Simulation	bitwise_not_vortex_buffer_mut[2048]	371.4 ns	429.7 ns	-13.57%
❌	Simulation	bitwise_not_vortex_buffer_mut[128]	246.1 ns	333.6 ns	-26.23%

---

_{Comparing ct/benchmarks-v3 (fbb0b39) with develop (d3ff1f1)}

Footnotes

1. 138 benchmarks were skipped, so the baseline results were used instead. If they were deleted from the codebase, click here and archive them to remove them from the performance reports. ↩