All Research Tracks
R18 · Developer Experience & Tooling

Reproducible Performance Regression Testing

SkeinDB's replay bundles enable reproducing correctness bugs, but performance issues are harder to reproduce. Performance depends on state not captured in traditional replay: compaction state, cache contents, concurrent load. Extending replay bundles to capture performance-relevant state enables reproducible performance regression testing — critical for database development confidence.

Research Proposal — Mapped to backlog in docs/RESEARCH_BACKLOG.md

🔬 What's Novel

🔧 Technical Approach

Phase 1 — State Identification

Identify performance-critical state beyond WAL: LSM level structure, block cache contents, connection pool state, and compaction queue depth. Determine minimal state for reproducibility.

Phase 2 — Bundle Extension

Extended replay bundle format with: LSM metadata snapshot, cache state approximation (LRU ordering, hot keys), and timing annotations on WAL records for pacing.

Phase 3 — Deterministic Replay

Reconstruct storage state from bundle, warm cache according to captured LRU ordering, and replay queries with timing annotations to reproduce concurrency patterns.

Phase 4 — Regression Detection

Capture bundles from production, replay in test environments with different code versions, compare latency distributions (p50, p99, p999), and alert on statistically significant deviations.

🧪 Hypotheses

H1

Performance-critical state (compaction level, cache hotness, concurrent operations) can be captured in extended replay bundles with manageable overhead.

H2

Deterministic replay reproduces performance characteristics within acceptable variance for regression detection.

H3

Performance replay enables root cause analysis for production performance regressions that are otherwise unreproducible.

🔗 SkeinDB Integration

Replay Bundles
Hash-Chained WAL
LSM / Compaction
Observability
Block Cache

📚 Key References