PostgreSQL Analytics: It's a Trap!
summary
This talk analyzes PostgreSQL's architecture to identify its real limitations for analytics (notably row-oriented storage) and explains how PostgreSQL 18 improves analytic workloads. It critiques common benchmarks (e.g., ClickBench) that favor ClickHouse or similar OLAP engines, compares trade-offs of abandoning ACID for speed, and argues that for most teams the best analytic engine is the one they already know how to operate and maintain.
AI-written summary · about
chapters (12)
- 0:00PostgreSQL analytics intro
- 2:52Switching analytics engine
- 3:13PostgreSQL architecture limits
- 6:23ClickHouse OLAP pros/cons
- 11:20DuckDB in-process limits
- 12:25Data modeling for analytics
- 18:54ACID vs OLAP tradeoffs
- 23:47Multi-engine costs
- 27:16Benchmarking pitfalls
- 35:44Engine switch strategy
- 40:40PostgreSQL default choice
- 41:01Q&A
related talks
49mCan PostgreSQL Compete with Analytics Specialists?
19mFrom OLTP to OLAP: Is PostgreSQL Eating Analytics Too?
34mAnalytics and AI with PostgreSQL - Mission Possible!
34mAnalytics and AI with PostgreSQL - Mission Possible!
21mAnalytics in Postgres –A decade in the making
35mBreakout Session
27mBuilding a PostgreSQL data warehouse
49mSELECT 'amazing features' FROM "postgresql"