docker-compose-hadoop.yml

version: '3.2'
services:
  namenode:
    image: bde2020/hadoop-namenode:2.0.0-hadoop2.7.4-java8
    networks:
      - network
    volumes:
      - namenode:/hadoop/dfs/name
    environment:
      - CLUSTER_NAME=test
    env_file:
      - ./hadoop.env
    ports:
      - target: 50070
        published: 50070
        protocol: tcp
        mode: host
      - target: 9000
        published: 9010
        protocol: tcp
        mode: host
    deploy:
      mode: replicated
      replicas: 1
      restart_policy:
        condition: on-failure
      # add placement constraint to allocate namenode on particular server
      placement:
        constraints:
          - node.role == manager

  datanode:
    image: bde2020/hadoop-datanode:2.0.0-hadoop2.7.4-java8
    networks:
      network:
        aliases:
         - datanode
    volumes:
      - datanode:/hadoop/dfs/data
    env_file:
      - ./hadoop.env
    environment:
      SERVICE_PRECONDITION: "namenode:50070"
    deploy:
      # mode global will deploy one datanode per docker swarm node
      mode: global
      restart_policy:
        condition: on-failure
      placement:
        constraints:
          - node.labels.type == hs-worker

volumes:
  datanode:
  namenode:

networks:
  network:
    external: true