Schedule - Nordic PGDay 2025
Postgres test data generation 101
Date: 2025-03-18
Time: 11:15–12:05
Room: Main Room
Level: Intermediate
The talk will start with the "why" side of testdata generation, as there are some non-obvious benefits of such exercises on top of the obvious ones. Also - where exactly to "plug in" this phase in the app developement lifecycle?
Next we cover in more depth the following most common Postgres test data generation aspects and techniques that everyone working more seriously with Postgres should be aware of, including:
- from generate_series() to LATERAL
- pgbench and building custom pgbench models
- how to get production-like data distributions without actually using any production data
- existing tooling like Synth & co
- mixing in AI and Python for more real-life looking data
We finish with tips for speeding things up along the way, and also listing some gotchas to look out for, as synthetic test data generation in reality also has some limitations and might not be a one-size-fits-all winner.
