One of the most important tasks of the query planner is to choose which relations to join, how to join them and in which order. How this works is however not widely known, even though the algorithm used is soon to be celebrating its 40th birthday.

This presentation will cover what join order selection is, how the algorithm works and how it's implemented in PostgreSQL, as well as look at what lies ahead in the future for join ordering in PostgreSQL. Hopefully it will demystify the query planner a little.