docker-compose.yml

version: '3.7'

services:
  datashare-worker:
    build:
      context: .
      target: worker
    deploy:
      mode: replicated
      replicas: 2
    depends_on:
      datashare_web:
        condition: service_started
    environment:
      DS_DOCKER_ML_LOG_LEVEL: DEBUG
      DS_DOCKER_ES_ADDRESS: http://elasticsearch:9200
      ICIJ_WORKER_TYPE: amqp
      ICIJ_WORKER_RABBITMQ_HOST: rabbitmq
      ICIJ_WORKER_RABBITMQ_PORT: 5672
    volumes:
      - type: bind
        source: ${PWD}/.data
        target: /home/user/src/app/.data

  # Adding rabbitmq to distribute Datashare tasks
  rabbitmq:
    image: rabbitmq:3.12.0-management
    # Set a host name to prevent rabbitmq to create new conf each time https://stackoverflow.com/a/53772874
    hostname: rmq
    container_name: datashare-rabbitmq
    healthcheck:
      test: rabbitmq-diagnostics -q status
      interval: 5s
      timeout: 2s
      retries: 10
      start_period: 5s
    ports:
      - "5672:5672"
      - "15672:15672"
    volumes:
      - type: volume
        source: rabbitmq-data
        target: /var/lib/rabbitmq


  # Unless commented, all lines below are copied from https://icij.gitbook.io/datashare/server-mode/install-with-docker
  datashare_web:
    # TODO: bump this to a version with serialization fixed
    image: icij/datashare:19.0.0
    hostname: datashare
    ports:
      - "8080:8080"
    environment:
      - DS_DOCKER_MOUNTED_DATA_DIR=${PWD}/data
    volumes:
      - type: bind
        source: ${PWD}/.data/datashare
        target: /home/datashare/Datashare
    depends_on:
      redis:
        condition: service_healthy
      elasticsearch:
        condition: service_healthy
      rabbitmq:
        condition: service_healthy
    # We need to switch the batchQueueType to AMQP and taskRoutingStrategy to Group compared to the default config
    command: >-
      --mode SERVER
      --batchQueueType AMQP
      --taskRoutingStrategy GROUP
      --dataDir /home/datashare/Datashare
      --pluginsDir /home/datashare/plugins
      --extensionsDir /home/datashare/extensions
      --authFilter org.icij.datashare.session.YesCookieAuthFilter
      --dataSourceUrl jdbc:postgresql://postgresql/datashare?user=datashare\&password=password 
      --defaultProject test-project
      --elasticsearchAddress http://elasticsearch:9200      
      --messageBusAddress amqp://guest:guest@rabbitmq:5672
      --queueType REDIS
      --redisAddress redis://redis:6379  
      --rootHost http://localhost:8080
      --sessionStoreType REDIS
      --sessionTtlSeconds 43200
      --tcpListenPort 8080

  elasticsearch:
    image: elasticsearch:7.17.25
    restart: on-failure
    volumes:
      - type: volume
        source: elasticsearch-data
        target: /usr/share/elasticsearch/data
        read_only: false
    environment:
      - "http.host=0.0.0.0"
      - "transport.host=0.0.0.0"
      - "cluster.name=datashare"
      - "discovery.type=single-node"
      - "discovery.zen.minimum_master_nodes=1"
      - "xpack.license.self_generated.type=basic"
      - "http.cors.enabled=true"
      - "http.cors.allow-origin=*"
      - "http.cors.allow-methods=OPTIONS, HEAD, GET, POST, PUT, DELETE"
    healthcheck:
      test: [ "CMD-SHELL", "curl --silent --fail elasticsearch:9200/_cluster/health || exit 1" ]
    ports:
      - "9200:9200"

  postgresql:
    image: postgres:12-alpine
    environment:
      - POSTGRES_USER=datashare
      - POSTGRES_PASSWORD=password
      - POSTGRES_DB=datashare
      # This is needed by the heathcheck command
      # @see https://stackoverflow.com/a/60194261
      - PGUSER=datashare
    volumes:
      - type: volume
        source: postgresql-data
        target: /var/lib/postgresql/data
    healthcheck:
      test: [ "CMD-SHELL", "pg_isready", "-U", "datashare", "-d", "datashare" ]
    ports:
      - "5432:5432"

  redis:
    image: ${REDIS_IMG}
    restart: on-failure
    volumes:
      - type: volume
        source: redis-data
        target: /data
    healthcheck:
      test: [ "CMD-SHELL", "redis-cli", "--raw", "incr", "ping" ]
    ports:
      - "6379:6379"

volumes:
  elasticsearch-data:
  postgresql-data:
  redis-data:
  rabbitmq-data: