The Lost Art of Database Design
summary
Argues that poor schema design is the top recurring problem and that small choices cascade into big costs: a user table keyed by email balloons storage and indexes, converting email to a CRC32 plus auto‑increment integer yields 3-4x space reduction and 2-3x query speedup. Notes Python MySQL drivers can show up to a 70% performance gap between versions. Covers sharding tradeoffs (Vitess/Citus/TiDB/Cockroach/Yugabyte), shard‑key pitfalls on aggregates, and the cloud’s “trust but verify” shared‑responsibility reality.
AI-written summary · about
chapters (12)
- 0:00Start
- 0:22Matt Yankovic Head of Open Source Strategy
- 1:24Designing for Success
- 6:23Stack Politics
- 15:17Capacity Planning
- 19:25Sharding and Clustering
- 20:54Difference between Clustering and Sharding
- 24:30Poor Database Design
- 25:15Why Do Data Types Matter
- 30:37Flexible Design
- 34:03Indexing
- 35:37Recapping
related talks
46mThe Lost Art of Database Design
01h 02mDesign Good Schemas - Get a Better Database
28mDungeons & Data: Bestiary of Antipatterns and How to Defeat Them
38mDubious Database Design
29mBuilding on Strengths, Learning from Differences
28mSmall Data: Databases in the real world
35m10 things a developer should know about databases
39m10 Things Developers should know about Databases