Skip to content

replit.md

Latest commit

 

History

History
68 lines (54 loc) · 3.08 KB

replit.md

File metadata and controls

68 lines (54 loc) · 3.08 KB

Workspace

Overview

OLAP Benchmark Sandbox — a research-grade benchmarking system that compares Postgres, DuckDB, and Spark on a 50M-row e-commerce dataset. Results are mapped to CMU 15-721 Advanced Database Systems concepts via an inline annotation layer (ConceptValidator).

pnpm workspace monorepo using TypeScript + Python.

Architecture

Python Benchmark Layer (olap-benchmark/)

  • data_generator.py — generates 50M-row Parquet dataset
  • loaders/load_all_systems.py — loads data into Postgres, DuckDB, Spark
  • benchmarks/benchmark_dashboards.py — Use Case 1: dashboard queries
  • benchmarks/benchmark_complex_joins.py — Use Case 2: 3-table join stress test
  • benchmarks/benchmark_variant_test.py — Use Case 3: VARIANT shredding acid test
  • benchmarks/benchmark_clustering.py — Use Case 4: clustering & zone map test
  • benchmarks/benchmark_acid_integrity.py — Use Case 5: ACID Integrity & Concurrency Control
  • benchmarks/benchmark_vectorized_execution.py — Use Case 6: Vectorized Execution & SIMD
  • utils/concept_validator.py — maps metrics → CMU 15-721 concept annotations
  • results/ — benchmark JSON output files (read by the API)

Express API Server (artifacts/api-server/)

  • GET /api/benchmarks/status — running state + available systems
  • GET /api/benchmarks/results/:useCase — reads results JSON from olap-benchmark/results/
  • POST /api/benchmarks/run/:useCase — spawns Python benchmark as child process
  • GET /api/benchmarks/logs/:useCase — SSE stream of live Python stdout

React Dashboard (artifacts/dashboard/)

  • Split-panel result cards: left=metrics chart, right=ConceptValidator annotation
  • Annotation panel: purple lecture badge, red smoking-gun box, blue interpretation, ✅/⚠️ status chip
  • SSE log streaming via EventSource
  • CMU 15-721 traceability matrix at bottom

Stack

  • Monorepo tool: pnpm workspaces
  • Node.js version: 24
  • Package manager: pnpm
  • TypeScript version: 5.9
  • API framework: Express 5
  • Validation: Zod, Orval (OpenAPI codegen)
  • Frontend: React + Vite + Tailwind + Recharts
  • Python: psycopg2, duckdb, pyspark, psutil

Key Commands

  • pnpm run typecheck — full typecheck across all packages
  • pnpm --filter @workspace/api-spec run codegen — regenerate API hooks and Zod schemas
  • pnpm --filter @workspace/api-server run dev — run API server locally

Running benchmarks (from olap-benchmark/ directory):

python3 benchmarks/benchmark_dashboards.py
python3 benchmarks/benchmark_complex_joins.py
python3 benchmarks/benchmark_variant_test.py
python3 benchmarks/benchmark_clustering.py

Result File Locations

  • olap-benchmark/results/use_case_1_dashboards.json
  • olap-benchmark/results/use_case_2_complex_joins.json
  • olap-benchmark/results/use_case_3_variant_acid_test.json
  • olap-benchmark/results/use_case_4_clustering.json
  • olap-benchmark/results/use_case_5_acid_integrity.json
  • olap-benchmark/results/use_case_6_vectorized_execution.json

See the pnpm-workspace skill for workspace structure, TypeScript setup, and package details.