Skip to content

Getting started

Getting started

The recommended way to get set up is with docker compose. At least some level of familiarity with docker compose and self-hosting is assumed.

This guide is for running on http://localhost:8000 behind a reverse proxy. However, it should be easy to adapt to serving the files directly with Caddy.

To get started there are 3 files needed:

docker-compose.yml

The main file to specify which docker containers to run. Once set up, this file shouldn't need to change for releases.

docker-compose.yml
name: numbat-tasks # (1)!

services:
    api:
        image: ghcr.io/hybridau/numbat-tasks-api:$VERSION
        env_file:
            - .env
        depends_on:
            database:
                condition: service_started
        volumes:
            - static_files:/static_files
        restart: always
    database:
        image: postgres:18-trixie
        volumes:
            - postgres_data:/var/lib/postgresql/
        env_file:
            - .env
        restart: always
    frontend: # (2)!
        image: ghcr.io/hybridau/numbat-tasks-frontend:$VERSION
        volumes:
          - static_files:/static_files
        env_file:
            - .env
    caddy:
        image: caddy:2-alpine
        ports: # (3)!
            - "127.0.0.1:8000:80" 
        volumes:
            - ./Caddyfile:/etc/caddy/Caddyfile:ro
            - caddy_data:/data
            - caddy_config:/config
            - static_files:/static_files:ro
        restart: always

volumes:
    postgres_data:
    caddy_data:
    caddy_config:
    static_files:
  1. This is used to keep all the containers and volumes in their own spaces, so you can have a different project with a volume called postgres_data and it won't conflict with this one.
  2. The frontend docker container will start, build the static files, write them out to /static_files/frontend and then exit. Caddy will continue to serve the files.
  3. To serve the site directly rather than behind a reverse proxy, change this to 
    ports:
  - "80:80" 
  - "443:443"

.env

All the settings should live in here. The following is the minimum configuration, but a list of all environment variables is available.

.env
# Can also use "latest" in place of a specfic version number
VERSION=0.14.2

# Reasonable defaults
POSTGRES_DB=numbat_tasks
POSTGRES_USER=numbat

# Must be changed for prod
ALLOWED_HOSTS=tasks.example.com
# echo POSTGRES_PASSWORD=$(dd if=/dev/urandom bs=1 count=18 status=none | base64) >> .env
# echo SECRET_KEY=$(dd if=/dev/urandom bs=1 count=42 status=none | base64) >> .env

Caddyfile

Caddyfile
:80 { # (1)!
    handle_path /api/static/* {
        root * /static_files/api
        file_server
    }
    handle /api/admin/* {  # (2)!
        # Block any IP address not on the allow list from accessing the Django admin console
        @blocked not remote_ip 10.1.1.16
        respond @blocked "Access Denied" 403
        reverse_proxy api:8000
    }
    handle /api/* {
        reverse_proxy api:8000
    }
    handle /assets/* {
        root * /static_files/frontend
        file_server
    }
    handle {
        root * /static_files/frontend
        file_server
        route {
            try_files {path} /index.html
            header /index.html Cache-Control "public, no-cache, max-age=0, must-revalidate"
        }
    }
}
  1. To serve the site directly, change this to 
    tasks.example.com {
   ...
    banana
  2. Restricting the Django admin console to specific IPs optional, but encouraged

Running

Place those three files in the same directory and to start the site run 

docker compose up -d

Create a user and login

See user management for steps on creating a user and logging in.

Delete everything and start again

Warning

This will permanently delete everything. All users, all lists, all tasks, everything.

docker compose down --volumes