Databases beyond the tutorials
summary
Advocates building faithful production-like test data for realistic capacity tests, splitting growth‑rates into separate schemas (financial, logs, PII) and rejoining with multi-source replication for a big reporting node. Deploy code that handles both old and new schemas before making schema changes, practice promoting replicas with Orchestrator, avoid enums, keep logs in UTC, ensure explicit primary keys to prevent replica table scans, and perform large writes in chunk updates of a couple thousand rows to avoid replica stalls.
AI-written summary · about
chapters (12)
- 0:00Introduction
- 0:52Why databases
- 1:50Testing
- 3:02Capacity testing
- 3:50Transaction vs analytic processing
- 6:33Data change
- 9:30System failures
- 11:06Dont be intimidated by SQL
- 12:52Object impedance mismatch
- 16:19Collectionbased APIs
- 21:49enum
- 22:31time zone
related talks
28mSmall Data: Databases in the real world
31mHelp! My Database Is The Bottleneck
47mEfficiency up, failed releases down; compliant production data for your test and dev environments
35mAvoiding Bad Database Surprises: Simulation and Scalability
26mNavigating Managed Databases for Over a Decade as a Non-DBA
11mThe Brave Little Database!
54mLessons learned from refactoring multi terabyte databases for the last 10 years
50mAdam Furmanek: You Must Test Your Databases And Here Is How