This project provides a recipe for building a basic Virtual Machine with Ubuntu 18.04 (Bionic), that sets up a simple Flask application under uwgi to be reverse-proxied by nginx, with nginx serving up the static parts of the application directly.
This is pretty much the configuration you'll need to deploy onto a Raspberry Pi under Raspbian Stretch, though there's nothing Pi-specific here.
You may have found your way here after struggling to get a similar stack deployed. You've probably seen stackoverflow posts about weird problems that can happen along the way if you don't get the incantations exactly right. Study the provisioning script carefully. If you must diverge, test frequently.
Building the VM requires VirtualBox and Vagrant. On Linux (Ubuntu)
sudo apt-get install virtualbox
gets you the former. http://vagrant.io gets you the latter.
Pull this down from github, then
vagrant up
On my Linux laptop, with an SSD over a mediocre network connection,
it takes about 10-15 minutes to build a VM.
Once done, http://localhost:8000 reaches the minimal flask starter
app running inside the VM.
vagrant ssh gets you a shell in the VM.
Exit that shell and vagrant suspend to put the VM to bed.
vagrant reload brings it back.
Those are the bare basics. Consult the Vagrant docs for more.
As you extend the Flask app, you can either check your work using flask
directly (see below), or you can restart services to see your changes.
For that, you'll want to vagrant ssh into the VM
and
sudo systemctl restart ngingx
sudo systemctl restart uwsgi
If you change something and http://localhost:8000/ fails to reach the app,
it's time to vagrant ssh into the VM.
First, check the Flask app.
cd /vagrant
FLASK_APP=starter/starter.py FLASK_DEBUG=1 venv/bin/flask run --host 0.0.0.0
From a browser outside of the VM, visit http://localhost:5000.
If there errors, Flask will spew errors into the browser, into the console, and sometimes both.
If your app works on :5000, the problem probably isn't your app.
To debug configuration configuration changes, try
cd /vagrant
uwsgi -H venv --http 0.0.0.0:5000 -M -w starter:app
then visit http://localhost:5000.
It's also useful to consult logs. The useful ones are
/var/log/uwsgi/emperor.log/var/log/uwsgi/starter.log/var/log/nginx/error.log/var/log/nginx/access.log
At the moment, it's a matter of pulling this code down onto a Pi,
ignoring Vagrantfile,
and manually performing the steps in provision/provision.sh.
Simplifying that is on my TODO list.
At the moment I'm taking the lazy approach of deploying to /vagrant on the Pi.
This is unwarrantied, use-at-your-own-risk-ware.
Pull requests that fix egregious problems are welcome;
feature PRs may be ignored.
And yeah, gunicorn is a viable alternative to uwsgi. I don't use it for reasons that might not be your reasons.