docker-compose.yml

version: "3.9"
name: analytics-engineer-assignment

x-airflow-common:
  &airflow-common
  image: airflow-clara-local:latest
  build:
    dockerfile: ./Dockerfile.airflow
    context: .
  env_file:
    - ./.env.airflow.local
    - ./.env.dbt.local
  environment:
    - DBT_POSTGRES_HOST=postgres
    - DBT_POSTGRES_PORT=5432
  volumes:
    - ./dags:/opt/airflow/dags
    - ./dbt:/opt/airflow/dbt:rw

x-airflow-depends-on:
  &airflow-depends-on
  depends_on:
    airflow-db-metadata:
      condition: service_healthy
    airflow-init:
      condition: service_completed_successfully

services:
  postgres:
    container_name: postgres
    hostname: postgres
    image: postgres:${POSTGRES_VERSION:-14.7}
    env_file:
      - ./.env.postgres.local
    ports:
      - ${POSTGRES_EXT_PORT:-5440}:${POSTGRES_PORT:-5432}
    restart: unless-stopped
    volumes:
      - ./scripts/postgres_init.sh:/docker-entrypoint-initdb.d/postgres_init.sh
      - postgres-volume:/var/lib/postgresql/data
    healthcheck:
      test:
        [
          "CMD",
          "pg_isready",
          "-U",
          "${POSTGRES_USER}",
          "-d",
          "${POSTGRES_DB}"
        ]
      interval: 10s
      timeout: 5s
      retries: 5

  bootstrap-dbt:
    image: bootstrap-dbt:latest
    build:
      dockerfile: ./Dockerfile.dbt
      context: .
    container_name: bootstrap-dbt
    depends_on:
      postgres:
        condition: service_healthy
    working_dir: /code
    entrypoint:
      - /bin/bash
      - -c
      - poetry run dbt seed --target prod --project-dir dbt --profiles-dir dbt
    restart: "no"
    env_file:
      - .env.dbt.local
    environment:
      - DBT_POSTGRES_HOST=postgres
      - DBT_POSTGRES_PORT=5432

  airflow-db-metadata:
    image: postgres:${AIRFLOW_POSTGRES_VERSION:-14.7}
    container_name: airflow-db-metadata
    hostname: airflow-db-metadata
    env_file:
      - ./.env.airflow.local
    environment:
      - POSTGRES_USER=${AIRFLOW_POSTGRES_USER}
      - POSTGRES_PASSWORD=${AIRFLOW_POSTGRES_PASSWORD}
      - POSTGRES_DB=${AIRFLOW_POSTGRES_DB}
      - POSTGRES_PORT=${AIRFLOW_POSTGRES_PORT}
    volumes:
      - airflow-db-metadata-volume:/var/lib/postgresql/data
    ports:
      - ${AIRFLOW_POSTGRES_PORT_HOST:-5441}:${AIRFLOW_POSTGRES_PORT:-5432}
    restart: on-failure
    healthcheck:
      test: [ "CMD", "pg_isready", "-U", "${AIRFLOW_POSTGRES_USER:-airflow}" ]
      interval: 5s
      retries: 5

  airflow-scheduler:
    <<: [*airflow-common, *airflow-depends-on]
    container_name: airflow-scheduler
    command: scheduler
    restart: on-failure
    ports:
      - 8793:8793

  airflow-webserver:
    <<: [*airflow-common, *airflow-depends-on]
    container_name: airflow-webserver
    restart: on-failure
    command: webserver
    ports:
      - ${AIRFLOW_WEBSERVER_PORT_HOST:-8080}:8080
    healthcheck:
      test:
        [
          "CMD",
          "curl",
          "--fail",
          "http://localhost:${AIRFLOW_WEBSERVER_PORT_HOST:-8080}/health"
        ]
      interval: 30s
      timeout: 30s
      retries: 5

  airflow-init:
    <<: *airflow-common
    container_name: airflow-init
    entrypoint: /bin/bash
    user: 0:0
    command:
      - -c
      - |
        mkdir -p /sources/logs /sources/dags /sources/plugins
        chown -R "${AIRFLOW_UID:-0}:0" /sources/{logs,dags,plugins}
        exec /entrypoint airflow db init

volumes:
  airflow-db-metadata-volume:
  postgres-volume:

networks:
  default:
    name: analytics-engineer-assignment