Squeezing Django performance for 14.9 million users on WhatsApp
summary
Built a WhatsApp service called ContactNDoH that handles 14.9 million users and 850 million messages, backs a health‑check risk algorithm and ships results into the national data lake. Uses Django + Postgres + Django REST Framework as a REST API and solves scale problems with cursor pagination, timestamp indexes, a custom admin pagination that falls back to Postgres approximate counts and statement timeouts, plus Locust and Faker for load tests, Django Debug Toolbar and EXPLAIN ANALYZE to find slow queries, and tuning (log_min_duration_statement, PyPy, gunicorn); system stays responsive in production.
AI-written summary · about
chapters (7)
- 0:00Introduction and background
- 4:23Contact NDOH project overview
- 7:57Database performance challenges
- 12:52API design with DRF
- 14:39Measuring and testing performance
- 21:46CPU and memory bottlenecks
- 23:54Q&A session
related talks
47mBuilding Scalable Web Apps
31mHigh Performance Django
47mReal World Django Q&A
42mA Nice Problem to Have: Django Under Heavy Load
21mHigh-Availability Django
39mHigh Performance Django: From Runserver to Reddit Hugs
28mThe life of a web request - techniques for measuring and improving Django application performance
47mHigh Performance Django at Ten: Old Tricks & New Picks