Skip to content

Latest commit

 

History

History
109 lines (76 loc) · 3.05 KB

getting_started.md

File metadata and controls

109 lines (76 loc) · 3.05 KB

You made it this far, you must be serious.

Ok, lets get started. You're gonna need the usual things to get started:

Requirements

  • Docker
  • Python 3.12
  • uv (for Python dependency management)
  • PostgreSQL
  • S3 Buckets (Provider of choice, we use Backblaze for production and S3Proxy1 for dev)
  • Domain with SSL certificate
  • make (seriously, you should have this)

Local development

  • PyCharm (not required but recommended)

Before you get Started

If you are on x86 then you might need to edit the docker-compose.yml.

Remove the platform: linux/arm64 from the s3proxy service.

Wait, I'm a lazy bastard and just want to see it work.

I got you, Bae. But you really need to get the OpenAI API key. When you have that come back.

Just run the following command:

make bootstrap-dev

It will :

  • pull containers
  • build the django container
  • install Python dependencies with uv
  • migrate the initial database
  • prompt you to create a superuser.

If you want to see all the things that does just peek at the Makefile.

NOTE: This does not start the crawlers. We will take that in the next step.

Post bootstrap steps

Now we are ready to rock. Let's spin up the full dev environment.

make dev-start

Note, the workers will also start, but they do nothing. You will need to activate them in the admin (See below).

When that is done point your browser to http://localhost:8000/admin, and you should see the application running. Login as the superuser.

Configure the site via the admin

  1. Site Name: Set up the name and domain in the admin.

  2. Global Site config Go to the "other" site configuration (yeah, yeah, I know) and check the following:

    1. Worker Enabled - This will enable the workers to run. Globally.
    2. Worker sleep seconds - This is the time in seconds that the workers will sleep between runs. 3JS Head: Javascript to run in the head of every page. This will be where you will put analytics, for example. 4JS Body: Javascript to run in the body of every page. This is the last tag in the body.

  3. Worker configs: Manage the finer-grained settings for workers:

    1. Is enabled: Enable the worker.
    2. Sleep seconds: The time in seconds that the worker will sleep between runs.
    3. Log Level: The log level for the worker. (This is important for debugging in production)

Production deployment

I currently use Dokku2 for deployment. It is a Heroku-like PaaS that you can run on your own servers. It is easy to use and has a lot of plugins.

Another option is Docker compose. You can use the docker-compose.yml file to run the application locally or on a server.

  • Dokku (it should also work with Heroku)
  • Docker Compose
  • PostgreSQL
  • Domain with SSL certificate

Footnotes

  1. https://github.com/gaul/s3proxy ↩

  2. https://dokku.com ↩