revolt-backend

Backend for radiorevolt.no

Environment settings

Settings are read from environment variables. By default settings are configured for development and should just work, but some changes are required in production.

Name	Default	Description
REVOLT_DEBUG	True	Enable Django debug. Set to false in production
REVOLT_ALLOWED_HOSTS	(Empty)	Hosts allowed to access webserver. e.g. 'radiorevolt.no'
DATABASE_URL	(BASE_DIR)/db.sqlite3	See dj-database-url
REVOLT_STATIC_ROOT	(BASE_DIR)/staticfiles	Folder where static files are located
REVOLT_MEDIA_ROOT	(BASE_DIR)/mediafiles	Files where uploaded media is located
REVOLT_SECRET_KEY	replace_this_secret_key	See Django SECRET_KEY. It is important that this is changed in production.
REVOLT_RAVEN_DSN	(Empty)	Sentry DSN using this format: https://user:pass@sentry.io/project

Testing and linting

The projects uses flake8, yapf, and isort to lint the project.

The easiest way to run the tests is to use Docker: docker-compose run test.
Alternatively use the docker script: ./docker/run_tests.sh.

Setup - Development

Docker

To start the project using Docker run docker-compose up dev which runs on http://localhost:8000
To update requirements run docker-compose up --build dev.
Migrations are automatically applied on start up.

No further changes are needed to run the backend for development.

Virtualenv

To run the project without using Docker virtualenv is recommended.

pip install virtualenv
virtualenv -p python3 venv
. venv/bin/activate

Install requirements

pip install -r requirements.txt
sudo apt-get install libpq-dev python3-dev
pip install psycopg2

Change settings

The default settings provided should work fine for development, but if any changes are needed a tool like direnv is recommended for handling environment variables.

Setup database and load data dump

python manage.py migrate
python manage.py collectstatic
python manage.py loaddata chimera_import_script/data.json

Start the development server

python manage.py runserver

Setup - Deployment

Docker

Docker can be used to run the backend in production. It is configured to use uwsgi on port 800. Staticfiles and mediafiles are served from local folders with the same name.

Start the backend using docker-compose up production.

Nginx

apt-get install nginx

Copy the nginx config (Updated as of 12/02/2017) to /etc/nginx/sites-available/radiorevolt.conf:

server {

  listen 80;
  # Type your domain name below
  server_name radiorevolt.no www.radiorevolt.no beta.radiorevolt.no;

  root /var/www;
  index index.html;
    
# Always serve index.html for any request
  location / {
    try_files $uri $uri/ /index.html;
    allow all;
  }
  location /static/ {
      alias   /var/www/static/;
  }
  
  location /media/ {
      alias   /var/www/media/;
  }
  location ~ /.well-known {
        allow all;
  }	
##
# If you want to use Node/Rails/etc. API server
# on the same port (443) config Nginx as a reverse proxy.
# For security reasons use a firewall like ufw in Ubuntu
# and deny port 3000/tcp.
##

   location /staging/api {
    proxy_pass http://localhost:9000;
    proxy_http_version 1.1;
    proxy_set_header Upgrade $http_upgrade;
    proxy_set_header Host $host;
    proxy_cache_bypass $http_upgrade;
  }

  location /graphql {
    proxy_pass http://localhost:8000/graphql;
    proxy_http_version 1.1;
    proxy_set_header Host $host;
  }

  location /admin/ {
    proxy_pass http://localhost:8000/admin/;
    proxy_http_version 1.1;
    proxy_set_header Host $host;
  }

  location /graphiql {
    proxy_pass http://localhost:8000/graphiql;
    proxy_http_version 1.1;
    proxy_set_header Host $host;
  }

  location /grappelli {
    proxy_pass http://localhost:8000/grappelli;
    proxy_http_version 1.1;
    proxy_set_header Host $host;
  }

  location /summernote/ {
    proxy_pass http://localhost:8000/summernote/;
    proxy_http_version 1.1;
    proxy_set_header Host $host;
  }
}

Enable the config by creating a symlink in /etc/nginx/sites-enabled:

sudo ln -s /etc/nginx/sites-available/radiorevolt.conf /etc/nginx/sites-enabled/radiorevolt.conf

If the server is only listening to IPv6 addresses replace this line:

listen 80;

With this line:

listen [::]:80 default ipv6only=on;

Clone repo

mkdir /webapps
cd /webapps
git clone https://github.com/Studentmediene/revolt-backend.git

Install PostgreSQL

sudo apt-get install postgresql postgresql-contrib
su - postgres
createuser --interactive -P

Enter name of role to add: revolt-backend
Enter password for new role: **********
Enter it again: (Ask your closest project leader for this)
Shall the new role be a superuser? (y/n) n
Shall the new role be allowed to create databases? (y/n) n
Shall the new role be allowed to create more new roles? (y/n) n

createdb --owner revolt-backend radiorevolt
\q
exit

Create an application user

groupadd --system webapps
useradd --system --gid webapps --shell /bin/bash --home /webapps/revolt-backend revoltbackend
chown -R revoltbackend: /webapps/revolt-backend

Virtualenvironment

cd /webapps/revolt-backend/
sudo apt-get install python3-pip
pip3 install virtualenv
virtualenv -p python3 venv
source venv/bin/activate

Install requirements

pip install -r requirements.txt
sudo apt-get install libpq-dev python3-dev
pip install psycopg2
pip install gunicorn
pip install setproctitle
sudo apt-get install supervisor

Copy data dump:

sudo cp chimera_import_script/data.json /webapps/revolt-backend/

Change settings

In production environment variables can be added to the supervisor/systemd service. See settings table at the top of the readme for full list.

Recommended changes:

REVOLT_DEBUG=false
REVOLT_ALLOWED_HOSTS="radiorevolt.no"
DATABASE_URL=postgres://revolt-backend:(password_here)@localhost/radiorevolt
REVOLT_STATIC_ROOT=/var/www/static/
REVOLT_MEDIA_ROOT=/var/www/media/
REVOLT_SECRET_KEY=(generate this with something like 'openssl rand -base64 32')

Setup database and load data dump

python manage.py migrate
python manage.py collectstatic
python manage.py loaddata data.json

Setup Gunicorn

Copy the Gunicorn config to /webapps/revolt-backend/gunicorn_start:

#!/bin/bash

NAME="revolt-backend"                               # Name of the application
DJANGODIR=/webapps/revolt-backend                   # Django project directory
SOCKFILE=/webapps/revolt-backend/run/gunicorn.sock  # we will communicate using this unix socket
USER=revoltbackend                                  # the user to run as
GROUP=webapps                                       # the group to run as
NUM_WORKERS=3                                       # how many worker processes should Gunicorn spawn
DJANGO_SETTINGS_MODULE=app.settings               # which settings file should Django use
DJANGO_WSGI_MODULE=app.wsgi                       # WSGI module name

echo "Starting $NAME as `whoami`"

# Activate the virtual environment
cd $DJANGODIR
source venv/bin/activate
export DJANGO_SETTINGS_MODULE=$DJANGO_SETTINGS_MODULE
export PYTHONPATH=$DJANGODIR:$PYTHONPATH

# Create the run directory if it doesn't exist
RUNDIR=$(dirname $SOCKFILE)
test -d $RUNDIR || mkdir -p $RUNDIR

# Start your Django Unicorn
# Programs meant to be run under supervisor should not daemonize themselves (do not use --daemon)
exec venv/bin/gunicorn ${DJANGO_WSGI_MODULE}:application \
  --name $NAME \
  --workers $NUM_WORKERS \
#  --user=$USER --group=$GROUP \
  --bind=unix:$SOCKFILE \
  --log-level=debug \
  --log-file=-

Make the file exectuable:

chmod +x /webapps/revolt-backend/gunicorn_start

Setup Supervisor

Copy the Supervisor config to /etc/supervisor/conf.d/revolt-backend.conf:

[program:revolt-backend]
command =sh /webapps/revolt-backend/gunicorn_start
user = revoltbackend
stdout_logfile = /webapps/revolt-backend/logs/gunicorn_supervisor.log
redirect_stderr = true
environment=LANG=en_US.UTF-8,LC_ALL=en_US.UTF-8

Add environment settings to the environment list.

Enable Supervisor:

sudo systemctl enable supervisor.service

Create log file and restart Supervisor:

mkdir -p /webapps/revolt-backend/logs/
touch /webapps/revolt-backend/logs/gunicorn_supervisor.log
supervisorctl reread

IF YOU GET AN ERROR (Supervisor/supervisor#121):

service supervisor start
supervisorctl reload
supervisorctl update