Skip to content

danjac/radiofeed-app

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

coverage

This is the source code for a simple, easy to use podcatcher web application. You are free to use this source to host the app yourself. A Dockerfile is provided for production deployments.

Development

Radiofeed requires the following basic dependencies to get started:

  • Python 3.12
  • uv

Note: if you don't have the right version of Python you can use uv python install 3.12.x.

For ease of local development a docker-compose.yml file is provided which includes Docker images:

  • PostgreSQL
  • Redis
  • Mailpit (for local email testing)

You can use these images if you want, or use a local install of PostgreSQL or Redis.

Current tested versions are PostgreSQL 16 and Redis 7.

The justfile has some convenient shortcuts for local development, including:

  • just install: download and install local dependencies
  • just update: update dependencies to latest available versions
  • just clean: remove all non-committed files and other artifacts
  • just serve: run the development server and Tailwind JIT compiler
  • just test: run the test suite

The install command will also create a .env file with default settings for local development, if one does not already exist.

Stack

The Radiofeed stack includes:

This stack was chosen for the following reasons:

  1. Locality of behavior: the behavior of a unit of code should be obvious from looking at the code.
  2. Performance: reduce the burden on end-user devices and bandwidth rampant with heavy SPA frameworks
  3. Batteries included: using popular, well-supported open source tools such as Django and PostgreSQL with a large number of features to avoid reinventing the wheel

Deployment

The following environment variables should be set in your production installation (changing radiofeed.app for your domain).

ALLOWED_HOSTS=radiofeed.app
DATABASE_URL=<database-url>
REDIS_URL=<redis-url>
ADMIN_URL=<admin-url>
[email protected]
EMAIL_HOST=mg.radiofeed.app
MAILGUN_API_KEY=<mailgun_api_key>
SECRET_KEY=<secret>
SENTRY_URL=<sentry-url>

Some settings such as DATABASE_URL may be set automatically by certain PAAS providers such as Heroku. Consult your provider documentation as required.

EMAIL_HOST should be set to your Mailgun sender domain along with MAILGUN_API_KEY if you are using Mailgun.

You should ensure the SECRET_KEY is sufficiently random: run the generate_secret_key custom Django command to create a suitable random string.

In some server configurations your load balancer (e.g. Nginx) may set the strict-transport-security headers by default. If not, you can set the environment variable USE_HSTS=true.

In production it's also a good idea to set ADMIN_URL to something other than the default admin/. Make sure it ends in a forward slash, e.g. some-random-path/.

A Dockerfile is provided for standard container deployments which should also work on Heroku or another PAAS.

Once you have access to the Django Admin, you should configure the default Site instance with the correct production name and domain.

The ansible directory contains full Playbooks for a multi-server deployment to a shared hosting provider such as Hetzner or Digital Ocean, this can be copied and modified for your particular use-case.

Deployment with Hetzner

A full deployment guide for Hetzner Cloud is available. You should have a DNS domain set up using e.g. Hetzner DNS or Cloudflare.

  1. Create a new Hetzner Cloud project
  2. Go to Security > API Tokens and create a new token with read/write access
  3. Copy the token and add it to your environment as HCLOUD_TOKEN
  4. Run terraform -chdir=tf init to download the Hetzner provider
  5. Run terraform -chdir=rf apply -var hcloud_token=$HCLOUD_TOKEN to create the necessary resources
  6. Change to the ansible directory and use ansible-vault create hosts to create a new inventory file and add the IP addresses of all the servers
  7. Do the same for vars/django.yml, vars/postgres.yml, and vars/site.yml to create the necessary variables (see example vars files for guidance)
  8. Run ansible-playbook -i hosts site.yml to deploy the application
  9. In your DNS configuration, point your domain to the new load balancer IP address

Scheduling background tasks

In production you should set up the following cron jobs to run these Django commands (with suggested schedules and arguments):

Parse podcast RSS feeds:

*/6 * * * * python manage.py parse_feeds

Generate similar recommendations for each podcast:

15 6 * * * python manage.py create_recommendations

Send podcast recommendations to users:

15 9 * * 1 python manage.py send_recommendations_emails

Note: ansible will set up these cron jobs for you if you use the provided Playbooks.