I have seen that everyone uses web pages with an NGINX or APACHE instead of serving the application directly with DJANGO, however I can not find the vulnerabilities that could be applied to attacks if you directly use the web with this Framework in the port 80 since all official documentation applies to this type of web server. If we mount a django application with uWSGI and NGINX it would be like this:
- the web client < - > the web server < - > the socket < - > uWSGI < - > Python
I understand that if we skip this:
- the web client < - > the socket < - > Python
It seems that with this last format we would lose many logs or traces that interest us but are there vulnerabilities about it?