One Billion Vectors On Postgres | Scaling Postgres 401

In this episode of Scaling Postgres, we discuss searching one billion vectors on Postgres, how to bypass a vector index search, unconventional optimizations and a 40x boost to listen/notify performance.

Content Discussed

• Scaling Vector Search to 1 Billion on PostgreSQL
• Retrieve and Rerank: Personalized Search Without Leaving Postgres
• Unconventional PostgreSQL Optimizations
• Turbocharging LISTEN/NOTIFY with 40x Boost
• The strange case of the underestimated Merge Join node
• Understanding ALTER TABLE Behavior on Partitioned Tables in PostgreSQL
• Introducing: Postgres Best Practices
• PostgreSQL on Kubernetes vs VMs: A Technical Decision Guide
• Postgres Serials Should be BIGINT (and How to Migrate)
• Who Contributed to PostgreSQL Development in 2025?
• Introducing pgEdge Load Generator: Realistic PostgreSQL Workload Simulation
• RegreSQL
• How I got started with DBtune (& why we chose Postgres) with Luigi Nardi
• Hacking Workshop for February 2026
• PostgreSQL: Transaction IDs and Virtual Transaction IDs
• The challenges of soft delete
• MERISE: The French Database Modeling Superpower That Could Save Your Data Model
• How Blocking-Lock Brownouts Can Escalate from Row-Level to Complete System Outages
• 4 causes of table bloat in PostgreSQL and how to address them
• PostgreSQL Contributor Story: Florin Irion
• The DATE Data Type in Oracle vs. PostgreSQL
• Exploration: CNPG Extensions(ImageVolume)
• Illinois Prairie PUG January Edition
• PDXPUG February 19th, 2026: What’s New in PostgreSQL 18
• Using the shared plan cache for Amazon Aurora PostgreSQL

YouTube Video

Podcast Audio