Pedro Müller

Passionate about clean code, modern web technologies and ai

less than 1 minute read

Deploy Django to Railway

Prep: requirements.txt, Procfile, env-based settings, static files handled (e.g. Whitenoise).

  • Procfile: web: gunicorn myproject.wsgi:application --bind 0.0.0.0:$PORT
  • Settings: DEBUG, SECRET_KEY, ALLOWED_HOSTS, DB from env. Use os.getenv(). For DB: PostgreSQL vars from Railway when you add a Postgres service.
  • Static: STATIC_ROOT, Whitenoise middleware, run collectstatic in build or release step.
  • Railway: Connect repo → add Postgres if needed → set env vars (including ones Railway gives you for DB) → deploy. They detect Python and build.

Env vars I set: DEBUG=False, SECRET_KEY, ALLOWED_HOSTS (e.g. .railway.app), plus DB vars if not auto-injected.

After deploy: Run migrations (Railway CLI or dashboard shell). Create superuser if needed. Check logs for 500s.

Gotchas: Static 404 → ensure collectstatic ran and Whitenoise is in middleware. DB connection → confirm env vars and that the DB service is linked. Restart after changing env vars.

Optional: health check endpoint (/health/ → 200), custom domain in Railway settings, Redis for cache if needed.