Time: 10:50 - 11:40
Room: Room 3
Managing and running PostgreSQL can pose interesting challenges when you’re operating databases at a large scale. This talk will cover some of the things I’ve encountered in the last year working with a 25TB+ PostgreSQL database. I will be using an unnamed real life client case-study as a guide for problems encountered and general solutions and guidelines.
Environment includes: 25TB database with significant growth 15 replicas WAL volume of 150GB/hr Transaction volume of 230M+/hr on primary alone Variable workload needs for different replicas
We will cover topics for DBA including: - Database tuning - WAL and replica management; backups strategies, WAL tools, bootstrapping new replicas - Issues encountered with such high transaction volumes, including MultiXact, vacuum tuning and challenges: what we had to change to get things working, etc - Indexing: identifying new indexes, creating them, and removing unused ones across the entire cluster - Growth projection/analysis: as the db gets big, how we project how/when to grow the database and the underlying infrastructure - Managing different workloads/workflows per server - Connection/user management