Django-Deployment-Guide

A simple guide for a quick django app deployment on any linux vps.

Steps:

Login To Your Server Over SSH:

This part is skipped because it doesn't belong to this context.

Install Server Dependencies

• Get the latest updates of the system packages.

foo@bar:~$ sudo apt-get update

• Upgrade the available upgradable packages.

foo@bar:~$ sudo apt-get -y upgrade

• Install nginx (proxy server, and assets server (css,js...))

foo@bar:~$ sudo apt-get -y install nginx

• Install supervisor (http://supervisord.org/)

foo@bar:~$ sudo apt-get -y install supervisor

• Install Python3.7

foo@bar:~$ sudo apt-get -y install python3.7

• Install pythoon virtual environment (will be used to isolate our project dependencies)

foo@bar:~$ sudo apt-get -y install virtualenv

Install any database server your project require, i won't cover this as it is out of context.

In case you are using MySQL (MariaDB), it requires some other dependencies to work.

foo@bar:~$ apt-get install python3.7-dev

foo@bar:~$ apt-get install python3.7-dev libmysqlclient-dev

foo@bar:~$ apt install build-essential

Configure The Project Environment:

• You should create a new system user with root privileges if you don't want to work with the root user, in my case i'm just going to use root.

• change your current directory to home and create a folder to deploy your app inside:

foo@bar:~$ cd /home

foo@bar:~$/home/ mkdir my_application_folder

foo@bar:~$/home/ cd my_application_folder

• now inside your application folder, pull your code from your repository (github or private server):

foo@bar:~$/home/my_application_folder git clone https://server.address/your/repo/path

• for the example the code root folder will be "application", now go inside it:

foo@bar:~$/home/my_application_folder cd application

• inside your application root folder (same level as manage.py file), we'll make the virtual environment to hold the project's dependencies:

application/
└── manage.py

foo@bar:~$/home/my_application_folder/application/ virtualenv --python=python3.7 venv

(you can name your environment as you like, but by convention we name "venv")

• Start using the virtual environment you have to activate it:

foo@bar:~$/home/my_application_folder/application/ source venv/bin/activate

you'll notice that the command line has changed, now it starts with "(venv)"

• install the project dependencies, by convention all the projects have a requirements file (check the requirements.txt for example), which holds the project's libraries' names and versions (optional) for the pip (package manager) to install, like so:

foo@bar:~$/home/my_application_folder/application/ pip install -r requirements.txt 

If you don't have a requirements.txt file (bad practice), you'll have to go over all the libraries manually like so:

foo@bar:~$/home/my_application_folder/application/ pip install lib_name 

• now you have to configure the project's settings, you'll do so by changing the settings.py file, use any editor you want,I personally use nano:

  • start by adding the server ip address and the host name (domain name) if provided to the allowed hosts property ex: ALLOWED_HOSTS=['192.168.1.1','www.mydomainname.com','mydomainname.com']
  • configure your database settings => out of context.
  • provide a STATIC_ROOT property if not yet provided, we'll create the respective folder later, this will be used to hold all the static files (js,css,images...), ex: STATIC_ROOT=os.path.join(BASE_DIR,'static_root')
  • save and exit.

• change directory to the application's root (same level as manage.py) and create the static root folder:

foo@bar:~$/home/my_application_folder/application/ mkdir static_root

• once again if you are runnning MySQL database you should add this dependency (the latest stable dependency for MySQL in linux):

foo@bar:~$/home/my_application_folder/application/ pip install mysqlclient==1.4.2.post1

• makemigrations if anything changed in the model layer, and migrate:

foo@bar:~$/home/my_application_folder/application/ python manage.py makemigration

foo@bar:~$/home/my_application_folder/application/ python manage.py migrate

• check for errors:

foo@bar:~$/home/my_application_folder/application/ python manage.py check

• check the django deployment check list (it will show some deployement requirements like https)

foo@bar:~$/home/my_application_folder/application/ python manage.py check --deploy

• test if your app configuration is ok:

foo@bar:~$/home/my_application_folder/application/ python manage.py runserver 0.0.0.0:8001

You can check if your app is running ok by going to your_ip_address:80001

if everything is ok go to the next step, otherwise check if you did eveyrhing good.

Configure Gunicorn:

for more details about gunicorn check https://gunicorn.org/

• install gunicorn in our virtual environement

foo@bar:~$/home/my_application_folder/application/ pip install gunicorn

• configure the gunicorn_start file to run our django app:

foo@bar:~$/home/my_application_folder/application/ cd venv/bin

foo@bar:~$/home/my_application_folder/application/venv/bin/ touch gunicorn_start

copy the content of gunicorn_start.txt file and make the appropriate changes according to your project

• make the gunicorn_start file executable

foo@bar:~$/home/my_application_folder/application/venv/bin/ chmod u+x gunicorn_start

• create the gunicorn socket folder:

foo@bar:~$/home/my_application_folder/application/venv/bin/ cd ../../../

change directory to the project folder (not the app folder) one level before manage.py inside the "my_application_folder".

foo@bar:~$/home/my_application_folder/ mkdir run

that's all for the gunicorn.

Configure Supervisor:

supervisor will be used to run your gunicorn server as a daemon (service), and keep your application up and running even after server crash or restart.

• change directory to supervisor configurations folder (this folder holds all the configurations for any daemon program):

foo@bar:~$/home/my_application_folder/ cd /etc/supervisor/conf.d/

foo@bar:~$/home/my_application_folder/etc/supervisor/conf.d/ touch anyname.conf

change the anyname as you want (your project name).

foo@bar:~$/home/my_application_folder/etc/supervisor/conf.d/ nano anyname.conf

copy the content of supervisor_config.txt and make the necessary changes according to your project.

foo@bar:~$/home/my_application_folder/etc/supervisor/conf.d/ cd /home/project my_application_folder/

foo@bar:~$/home/my_application_folder/ mkdir logs

foo@bar:~$/home/my_application_folder/ cd logs

foo@bar:~$/home/my_application_folder/logs touch gunicorn-error.log

foo@bar:~$/home/my_application_folder/logs touch gunicorn-out.log

• now you'll have to enable and start the supervisor service:

foo@bar:~$/home/my_application_folder/logs systemctl enable supervisor

foo@bar:~$/home/my_application_folder/logs systemctl start supervisor

• now reread new scripts and update the supervisor registery:

foo@bar:~$/home/my_application_folder/logs supervisorctl reread

foo@bar:~$/home/my_application_folder/logs supervisorctl update

• now you should see that your program name appearing and supervisor will tell you that it is added.

• you can check your program status by:

foo@bar:~/home/my_application_folder/logs$ sudo supervisorctl status program_name

you should see the programe name and "RUNNING" next to it with pid and uptime.

now your app is up and running on the gunicorn, but it can't be accessed from outside over http, that's where nginx comes.

Configure Nginx:

nginx will be used as a proxy server to redirect the http requests to gunicorn server, and also serve the media and static files, for bigger apps you should consider making a different server to serve the media files.

• create the site config file:

foo@bar:~$ cd /etc/nginx/sites-available/
foo@bar:~/etc/nginx/sites-available/$ touch project_name (or any name you want)
foo@bar:~/etc/nginx/sites-available/$ nano project_name (copy the content of nginx_config.txt and make the necessary changes according to your project)
foo@bar:~/etc/nginx/sites-available/$ cd /home/my_application_folder/logs/ (go to the project folder and make the nginx logs files)
foo@bar:~/home/my_application_folder/logs/$ touch nginx-access.log
foo@bar:~/home/my_application_folder/logs/$ touch nginx-error.log

• create a symbolic link to enable the website.

    foo@bar:~/home/my_application_folder/logs/$ sudo ln -s /etc/nginx/sites-available/config_file_name /etc/nginx/sites-enabled/config_file_name 

• delete the nginx default website:

    foo@bar:~/home/my_application_folder/logs/$ rm /etc/nginx/sites-enabled/default

• restart the nginx server service:

    foo@bar:~/home/my_application_folder/logs/$ service nginx restart

now if you go to your domain name you can access your application, but you'll notice that the static files are missing,no css,js... *that because we didn't collect the static files into the static_root folder, so: * change directory to your application, same level as manage.py console foo@bar:~/path/to/app/$ python manage.py collectstatic
check your website now, it should be up and running perfectly. note that if you make any changes to your code, you'll have to restart the supervisor daemon only:

    foo@bar:~$ supervisorctl restart program_name

Done.