This is the example code for the book The Temple of Django Database Performance by Andrew Brookins.
This code is for version 2 of the book, published August 2020. All buyers of the version 1 ebook (2019) have access to version 2 as a free download.
If you purchased the 2019 edition of the print book, this code is not substantially different than the code referenced in that book. However, this code includes a new example on using materialized views.
This project uses Docker to set up its environment, and it includes a Makefile to let you run docker-compose commands more easily.
Run make build to build the images for the environment.
You'll also want to run docker-compose run web ./manage.py createsuperuser to create a superuser for yourself.
Run make dev to run Redis, Postgres, and the Django web application. The example's servers bind their ports to localhost, so you can visit the app at https://localhost:8000 once it's running.
Run docker-compose logs web to view logs for the web application. Likewise, 'postgres' and 'redis' will show logs for those servers.
Run make test to run the tests. Tests run in a container. If you drop in "import ipdb; ipdb.set_trace()" anywhere in the project code, you'll drop into a debugging session if the tests hit that code.
Recreating many of the performance problems in this book requires a large amount of data in your database. This project includes a management command that will generate analytics events sufficient to cause performance problems.
Here's an example of using the management command to generate 500,000 analytics events for the user with ID 1 (in my case, this is my admin user):
$ docker-compose run web ./manage.py generate_events --num 500000 --user-id 1
This example code is copyright 2020 Andrew Brookins.