Thomas Munro

PostgreSQL added single-node parallel query execution in a series of leaps and bounds a few years ago, but it is far from finished. This talk is a reflection on design choices made, path dependencies that influenced those choices, where they sit in the design space, and a selection of major opportunities that remain open for future development.

This is a three part talk. The first part gives a very high level overview of the areas of development in recent major releases, including planning, execution, partitioning, replication, backups, monitoring, security, I/O, ongoing modernisation work and more.

Traditionally, PostgreSQL did all relation I/O as sequential 8KB pread() or pwrite() system calls. This talk is about a new "streaming" programming model that allows multiple blocks to be read or written at the same time, and eventually in the background ahead of time.

A look at PostgreSQL transaction logging in historical POSTGRES and how it evolved in modern PostgreSQL, how the approach compares to other systems, and how it might evolve in future releases.

An overview of how parallel query execution works, with a look at some of the gory details that let you use multiple CPU cores for a single query in recent PostgreSQL releases.