Introduction to Fair-Use TPC Benchmarking Kits
summary
Overview of freely available TPC fair‑use kits that run on Linux and OSDL test kits used to generate and load data then run tests; demos use TPC‑C, TPC‑E, TPC‑H and DS workloads plus mentions of BenchBase and HammerDB. Key specifics: primary metrics (tpmC, queries/hour), response‑time breakdowns including 90th percentiles and rollbacks (TPC‑C: ~1% rollback on New‑Order), storage scaling rules (1,000,000 tpmC = ~7 TB raw & ~770k emulated users), checkpointing rules (checkpoints every 30 minutes), and fair‑use limits on public comparisons.
AI-written summary · about
chapters (15)
- 0:00Introduction to TPC benchmarks
- 1:00What is a TPC benchmark
- 4:16TPC organization and resources
- 5:41Understanding TPC results
- 12:13Fair use of TPC materials
- 16:44TPC-C benchmark overview
- 27:19Running TPC-C tests
- 32:15Evaluating TPC-C performance
- 36:51Customizing TPC-C workloads
- 38:17TPC-E benchmark overview
- 41:47Running TPC-E tests
- 44:34TPC-E workload adjustments
- 46:20TPC-H benchmark overview
- 48:28Running TPC-H tests
- 50:15Query-level testing with TPC-H
related talks
44mDemystifying Benchmarks: How to Use Them To Better Evaluate Databases
52mHow to use the DBT 3 test kit
29mPerformance Archaeology - 20 years of improvements
28mTPC-C Benchmark: Scaling YugabyteDB to 150,000 Warehouses
06mLightning Talk: Performance Testing of FoundationDB using TPC-C
37mContinuous Performance Benchmarking for Vitess
44mBenchmarking your commits: The Postgres Performance Farm
31mDemanding the impossible rigorous database benchm