Coding for Entrepreneurs

Hello Linux: Gunicorn & Supervisor

By Justin

Hello Linux: Gunicorn & Supervisor
This is the fourth post of a many part series. This post is the starter post for the whole series.
Guincorn is our production WSGI server for running Django or Python web apps in a production environment.
Supervisor is a process manager to ensure Gunicorn (among other things) starts, restarts, and runs when we need it to.
Requirements
Be sure to complete
Our Live Server
  • Ubuntu 18.04

1. SSH into your server

ssh [email protected]
Replace [email protected] with your user / ip

2. Install & Start Supervisor

sudo apt-get update -y

sudo apt-get install supervisor -y 

sudo service supervisor start

3. Install Gunicorn in our Working Directory's Virtualenv

Find the Virtualenv bin that you created in this post
Mine is /var/www/myproject/bin/
/var/www/myproject/bin/python -m pip install gunicorn

4. Verify Gunicorn works

/var/www/myproject/bin/gunicorn --workers 3
You should see
usage: gunicorn [OPTIONS] [APP_MODULE]
gunicorn: error: No application module specified.
You should not see:
-bash: /var/www/myproject/bin/gunicorn: No such file or directory
So how many workers should we use? A rule of thumb is to do (2 * num_cores) + 1
How do you get the number of cores your Linux system has?
grep -c ^processor /proc/cpuinfo
Let's say that returns 2, you'd have 5 workers (2 * 2) = 4 + 1 = 5

5. Create a Supervisor Process

With supervisor, we can run step 4 automatically, restart it if it fails, create logs for it, and start/stop it easily.
Basically, this ensures that our web server will continue to run if you push new code, server reboots/restarts/goes down and back up, etc.
Of course, if a catastrophic error occurs (or if bad code is in your Django project) then this process might fail as well.
All supervisor processes go in:
/etc/supervisor/conf.d/
So, if you ever need to add a new process, you'll just add it there.
Let's create our project's gunicorn configuration file for supervisor.
touch /etc/supervisor/conf.d/myproject-gunicorn.conf
Now you should see:
$ ls -al /etc/supervisor/conf.d/
myproject-gunicorn.conf
Now, let's add the base settings:
[program:myproject_gunicorn]
user=root
directory=/var/www/myproject/src/
command=/var/www/myproject/bin/gunicorn --workers 3 --bind unix:myproject.sock cfehome.wsgi:application
 
autostart=true
autorestart=true
stdout_logfile=/var/log/myproject/gunicorn.log
stderr_logfile=/var/log/myproject/gunicorn.err.log
Let's break it down line by line.
[program:myproject_gunicorn]
: the myproject_gunicorn is the name of the supervisor process. Naturally, myproject_gunicorn is an arbitrary name. Once we complete Step 5, you'll be able to run the following commands: - sudo supervisorctl status myproject_gunicorn - sudo supervisorctl start myproject_gunicorn - sudo supervisorctl stop myproject_gunicorn - sudo supervisorctl restart myproject_gunicorn
user=root
: Set the user you want to allow access to this process. This must be set.
directory=/var/www/myproject/src/
This is the working directory for your project (notice mine has src at the end)
command=...
This is the command you want your process to run. Our command is: - /var/www/myproject/bin/gunicorn This is the recently installed gunicorn bash command - --workers 3 This is a gunicorn argument for the number of workers - --bind unix:myproject.sock This creates a socket for your gunicorn to bind to within your directory you set above. - cfehome.wsgi:application Where is your actual application? Django projects typically have a wsgi.py file that declares an application. This is the path to that exact declaration. In my case it's src directory -> cfehome module -> wsgi module -> application
autostart
Do you want this process to auto start? We do.
autorestart
Do you want this process to auto restart if things fail? We do.
stdout_logfile
and stderr_logfile are locations for log files. You might need to run mkdir /var/log/myproject/ to ensure these files are actually added. These log files are great for diagnosing errors.

6. Update Supervisor

supervisorctl reread
supervisorctl update

7. Check Our Supervisor Program/Process Status

As we mentioned when we created the myproject_gunicorn supervisor program, we can now do:
sudo supervisorctl status myproject_gunicorn
A few other useful commands (again):
  • sudo supervisorctl start myproject_gunicorn
  • sudo supervisorctl stop myproject_gunicorn
  • sudo supervisorctl restart myproject_gunicorn

Wrap Up

We now have
Now we need to
Discover Posts