Deploying Django Project Using both ASGI and WSGI
This deployment is based on the experiences from the previous two posts.
Deploying Async LLM-Chat with Message History
Integrating LLM Using API to a Django and React App
config/settings.py
# config/settings.py
WSGI_APPLICATION = 'config.wsgi.application'
ASGI_APPLICATION = 'config.asgi.application'
Procfile
In the Procfile, you should add and parameterize both of the servers: My choices are Gunicorn for WSGI and Daphne for ASGI.
web: gunicorn config.wsgi:application -b 0.0.0.0 & daphne -b 0.0.0.0 -p $PORT config.asgi:application
react/views.py
Set also HTTP/HTTPS protocol in IndexView.
# react/views.py
from django.views.generic import TemplateView
class IndexView(TemplateView):
template_name = 'index.html'
def get_context_data(self, **kwargs):
context = super().get_context_data(**kwargs)
# Check for forwarded protocol (for proxies like Railway)
forwarded_proto = self.request.META.get('HTTP_X_FORWARDED_PROTO')
is_secure = self.request.is_secure() or forwarded_proto == 'https'
# Set HTTP/HTTPS protocol
http_protocol = 'https://' if is_secure else 'http://'
context['BACKEND_URL'] = f"{http_protocol}{self.request.get_host()}"
# Set WS/WSS protocol
ws_protocol = 'wss://' if is_secure else 'ws://'
context['WS_URL'] = f"{ws_protocol}{self.request.get_host()}"
return context
Update files in frontend\
Add the BACKEND_URL|safe <script> to the <head> tag in frontend/index.html and call it with windows.BACKEND_URL in axios.jsx and wherever necessary.
Add the WS_URL|safe <script> to the <head> tag in frontend/index.html and call it with windows.WS_URL in Chat.jsx.