Deployment

For production, compile your front-end (or use the shipped bundle), collect the static files, run with DEBUG=False, and serve through a WSGI/ASGI server. The package supports a fully Node-free deploy as well as the Vite pipeline.

Option A — Node-free (static bundle)

The package ships a compiled asset bundle inside static/adminlte/dist/, so you can deploy with zero Node. Switch the front-end to it and collect:

# settings.py
ADMINLTE = {"assets_mode": "static"}

python manage.py collectstatic --noinput
gunicorn config.wsgi

With "static", django-vite is never imported, so you don't even need it installed.

Option B — Vite pipeline

Keep assets_mode="vite" to serve your own compiled bundle (and the optional lazy-loaded plugins). Build, then collect:

npm run build                                # writes assets/dist/ + manifest.json
python manage.py collectstatic --noinput
gunicorn config.wsgi

Make sure django-vite runs in production (non-dev) mode so it reads the hashed manifest instead of the HMR client:

DJANGO_VITE = {
    "default": {
        "dev_mode": DEBUG,                   # False in production
        "manifest_path": BASE_DIR / "assets" / "dist" / "manifest.json",
    }
}

Production checklist

Static files with WhiteNoise

The demo project is a twelve-factor starter that reads everything environment-specific from the environment, with production-safe defaults. With DEBUG=False it automatically enables HSTS, SSL redirect, secure cookies, and WhiteNoise's compressed manifest static storage. A typical deploy is:

# Environment: SECRET_KEY, DEBUG=False, ALLOWED_HOSTS, DATABASE_URL=postgres://…
npm run build                                # (Option B only)
python manage.py migrate
python manage.py collectstatic --noinput
gunicorn config.wsgi