PostgreSQL Observed—and Explained
summary
Never disable autovacuum — it prevents xid wraparound and you can set log_autovacuum_min_duration=0 to see detailed runs; autovacuum defaults to three workers and is tunable. Keep plenty of free disk (98% full triggers panic) and ensure WAL archiving actually copies logs off-server. Take multiple backups and practice restores. Use EXPLAIN ANALYZE to find queries that drop from minutes to milliseconds; prefer a pool over raising max_connections=100 into the thousands.
AI-written summary · about
chapters (14)
- 0:00Introduction and speakers
- 1:11Why this talk
- 3:22Autovacuum explained
- 7:19Disk space management
- 9:47Importance of backups
- 11:10Foreign data wrappers
- 12:56Primary keys and indexes
- 18:38Testing upgrades first
- 20:05Minor version upgrades
- 22:19Query performance issues
- 26:10Database connections
- 27:17Reading the docs
- 29:26Asking questions
- 32:04Q&A and feedback
related talks
32mOptimizing autovacuum: PostgreSQL’s vacuum cleaner
27mMaintaining Large Tables in PostgreSQL
47mAutoVacuum - Tuning and Monitoring
25mScaling PostgreSQL Beyond Query Optimization
55mUnderstanding Autovacuum
31mOptimizing Postgres for write heavy workloads ft. Checkpoint and WAL configs
49mUnderstanding and Fixing Autovacuum
33mPostgres Storytelling: Support in the Darkest Hour