Deploying Flask application with uwsgi and nginx

Question

I am trying to deploy a simple web app that I have built using Python and Flask.

My app has the following structure:

/var/www/watchgallery/
    + app
        + __init__.py
        + views.py
        + templates
    + flask #virtual environment for Flask
    + run.py #script I used in my machine to start the development Flask server
    + watchgallery_nginx.conf
    + watchgallery_uwsgi.ini
    + watchgallery_uwsgi.sock

For this purpose of deploying, I am following this link: http://vladikk.com/2013/09/12/serving-flask-with-nginx-on-ubuntu/

In this tutorial, the Flask app consists of only a hello.py file. The way he configures his uwsgi file is like this (/var/www/demoapp/demoapp_uwsgi.ini):

[uwsgi]
#application's base folder
base = /var/www/demoapp

#python module to import
app = hello
module = %(app)

home = %(base)/venv
pythonpath = %(base)

#socket file's location
socket = /var/www/demoapp/%n.sock

#permissions for the socket file
chmod-socket    = 666

#the variable that holds a flask application inside the module imported at line #6
callable = app

#location of log files
logto = /var/log/uwsgi/%n.log

I have tried to apply the same logic to my uwsgi.ini file, but I am doing something wrong. This is how my file looks like:

[uwsgi]
#application's base folder
base = /var/www/watchgallery

#python module to import
app = run
module = %(app)

home = %(base)/flask
pythonpath = %(base)

#socket file's location
socket = /var/www/watchgallery/%n.sock

#permissions for the socket file
chmod-socket    = 666

#the variable that holds a flask application inside the module imported at line #6
callable = app

When I am developing my app in my local machine, I run this command to start de the server: ./run.py.

This is my run.py file:

#!flask/bin/python
from app import app
app.run(debug = False)

Now, my question is: how should my uwsgi.ini file look like given that my Flask app consists of more than a single file?

Answer 1

It doesn't matter how complex your application is. You tell uWSGI where the entry is, the rest is processed normally with Python imports.

In your case the entry is module = %(app) and callable = app. So uWSGI will load the module and send requests to the callable which is a Flask application.

Now since the requests are to be served by uWSGI and not Flask's server, you don't need the app.run(debug = False) line. But you can keep development and production code the same with this trick:

#!flask/bin/python

from app import app

if __name__ == "__main__":
    app.run(debug = False)