Docker compose for Phoenix project

Line by line explanation of how to write a docker compose file for phoenix project for local development and testing. Prepare the application for release by following official documentation

docker compose is primarily meant for local development and testing although as per the official documentation there is plan to create prodcution ready version. An environment created by docker compose very much resembels a k8s cluster and behaves the same way in terms of deployment, scaling and networking.

version is an informative tag used to specify the schema version supported by the file. docker doesn't use this to validate the file syntax, this is an optional parameter included for backward compatibility

name specify the project name using this top level config. The project can be included in service environemnt using variable COMPOSE_PROJECT_NAME

services defines a list of container that are supposed to run as part of compose deployment. For a full stack phoenix application, one service is defined to run the phoenix container and another to run the database.

# name of the service
codejam_phoenix:
    # name or tag for the image of the service
    image: codejam/phoenix
    # build steps. for development purposes build the image form the
    # local codebase.
    build:
      # context points to the base path of phoenix application code
      # depending on the directory structure the it could be . or relative path
      context: ./codejam
      # name of the dockerfile if it is not the default name
      dockerfile: codejam.dev.Dockerfile
    # name of the container that you can look up in docker dashboard or CLI output
    container_name: codejam_phoenix
    # environment variables required to run the application
    env_file: ./codejam/dev/.env
    # additional environment variable with values passed inline
    environment:
      APP_ENV: development
    # forward the application port 4000 to host machine
    ports:
      - "4000:4000"
    volumes:
      - codejam-data:/tmp/codejam-data
    networks:
      - codejam_network
    # make sures that the container will be restarted in case of an error
    # but will stay stopped if explicitly killed
    restart: unless-stopped
    # name service on which this application should depend and wait on
    depends_on:
      codejam_postgres:
        condition: service_healthy
        restart: true
      codejam_minio:
        condition: service_started
    healthcheck:
      test: ["CMD", "curl", "-f", "http://localhost"]
      interval: 1m30s
      timeout: 10s
      retries: 3

  # postgres database for the application
  codejam_postgres:
    # pull the image from repository instead of building
    image: postgres:16.2
    container_name: codejam_postgres
    environment:
      # for local development run db in trust mode and without password
      - POSTGRES_HOST_AUTH_METHOD=trust
      # define the name of the database
      - POSTGRES_DB=codejamdb
    ports:
      - "7432:5432"
    networks:
      - codejam_network
    restart: unless-stopped
    # health check lets compose know that service is up
    # and it is safe to start dependant services
    healthcheck:
      test: ["CMD", "pg_isready"]
      interval: 1m30s
      timeout: 10s
      retries: 3

networks

define internal network that will be used by the containers in deployment to discover and connect to other services

networks:
  codejam_network:
    driver: bridge

volumes

persistent data volumes for the services, if required

volumes:
  codejam-data:
  codejam-minio-data: