To get the show notes as well as get notified of new episodes, visit:
https://www.scalingpostgres.com/episodes/424-too-many-table-danger/

In this episode of Scaling Postgres, we discuss the dangers of too many tables, don't worry about extending checkpoint_timeout, a disaster recovery process, and non-repeatable sums with floating point numbers.

Want to learn more about Postgres performance? 
Join my FREE training called Postgres Performance Demystified here: https://www.scalingpostgres.com/courses/postgres-performance-demystified/

Timestamps:

#postgres #postgresql

Scaling Postgres

Episode 424 of Scaling Postgres covers several PostgreSQL topics: the memory problems caused by having tens of thousands of tables (each backend accumulating ~500MB in cache memory context), checkpoint_timeout tuning debates (15-30 minutes recommended vs. the common 5-minute default), disaster recovery runbook best practices, floating-point arithmetic pitfalls when summing monetary values, a WAL replay self-deadlock bug in large database recovery, a PostgreSQL checkpoint infinite retry loop bug, the new PG Health Check open-source diagnostic tool, sequence synchronization for logical replication upgrades in PostgreSQL 19, using PostgreSQL as a temporal database, and the upcoming SQL 2025 standard features.

Too Many Table Danger | Scaling Postgres 424