Running the Deployment

To run the containers there are a couple of options but primarily the two main ways are with either Docker Compose or Kubernetes (via Helm), the gaffer-docker repository has some examples of how to run both.

Docker Compose

For this example we will use a slightly modified version of the docker compose config file used in the repository.

docker-compose.yaml

version: "3.7"

services:

  zookeeper:
    image: zookeeper:${ZOOKEEPER_VERSION}
    healthcheck:
      test: echo ruok | nc 127.0.0.1 2181 | grep imok
      interval: 30s
      timeout: 5s
      retries: 3
    container_name: zookeeper
    hostname: zookeeper
    environment:
    - ZOO_SERVERS=server.1=zookeeper:2888:3888;2181
    - ZOO_4LW_COMMANDS_WHITELIST=*
    volumes:
    - /data
    - /datalog

  hdfs-namenode:
    image: gchq/hdfs:${HADOOP_VERSION}
    depends_on:
      zookeeper:
        condition: service_healthy
    healthcheck:
      test: curl -f http://localhost:9870 || exit 1
      interval: 30s
      timeout: 10s
      retries: 3
    command: namenode
    container_name: hdfs-namenode
    hostname: hdfs-namenode
    environment:
    - HADOOP_CONF_DIR=${HADOOP_CONF_DIR}
    ports:
    - 9870:9870
    volumes:
    - ./configs/hdfs:${HADOOP_CONF_DIR}:ro
    - /var/log/hadoop
    - /data1
    - /data2

  hdfs-datanode:
    image: gchq/hdfs:${HADOOP_VERSION}
    depends_on:
      hdfs-namenode:
        condition: service_healthy
    command: datanode
    container_name: hdfs-datanode
    hostname: hdfs-datanode
    environment:
    - HADOOP_CONF_DIR=${HADOOP_CONF_DIR}
    volumes:
    - ./configs/hdfs:${HADOOP_CONF_DIR}:ro
    - /var/log/hadoop
    - /data1
    - /data2

  accumulo-master:
    image: gchq/gaffer:${GAFFER_VERSION}-accumulo-${ACCUMULO_VERSION}
    depends_on:
      hdfs-namenode:
        condition: service_healthy
    healthcheck:
      test: cat /proc/net/tcp | grep 270F
      interval: 30s
      timeout: 5s
      retries: 3
      start_period: 10s
    build:
      context: .
      args:
        GAFFER_VERSION: ${GAFFER_VERSION}
        BASE_IMAGE_NAME: gchq/accumulo
        BASE_IMAGE_TAG: ${ACCUMULO_VERSION}
    command: master
    container_name: accumulo-master
    hostname: accumulo-master
    environment:
    - ACCUMULO_CONF_DIR=${ACCUMULO_CONF_DIR}
    - HADOOP_USER_NAME=hadoop
    volumes:
    - ./configs/accumulo:${ACCUMULO_CONF_DIR}:ro
    - /var/log/accumulo

  accumulo-tserver:
    image: gchq/gaffer:${GAFFER_VERSION}-accumulo-${ACCUMULO_VERSION}
    depends_on:
      accumulo-master:
        condition: service_healthy
    healthcheck:
      test: cat /proc/net/tcp | grep 270D
      interval: 30s
      timeout: 5s
      retries: 3
    command: tserver
    container_name: accumulo-tserver
    hostname: accumulo-tserver
    environment:
    - ACCUMULO_CONF_DIR=${ACCUMULO_CONF_DIR}
    - HADOOP_USER_NAME=hadoop
    volumes:
    - ./configs/accumulo:${ACCUMULO_CONF_DIR}:ro
    - /var/log/accumulo

  accumulo-monitor:
    image: gchq/gaffer:${GAFFER_VERSION}-accumulo-${ACCUMULO_VERSION}
    depends_on:
      accumulo-master:
        condition: service_healthy
    command: monitor
    container_name: accumulo-monitor
    hostname: accumulo-monitor
    environment:
    - ACCUMULO_CONF_DIR=${ACCUMULO_CONF_DIR}
    - HADOOP_USER_NAME=hadoop
    ports:
    - 9995:9995
    volumes:
    - ./configs/accumulo:${ACCUMULO_CONF_DIR}:ro
    - /var/log/accumulo

  accumulo-gc:
    image: gchq/gaffer:${GAFFER_VERSION}-accumulo-${ACCUMULO_VERSION}
    depends_on:
      accumulo-master:
        condition: service_healthy
    command: gc
    container_name: accumulo-gc
    hostname: accumulo-gc
    environment:
    - ACCUMULO_CONF_DIR=${ACCUMULO_CONF_DIR}
    - HADOOP_USER_NAME=hadoop
    volumes:
    - ./configs/accumulo:${ACCUMULO_CONF_DIR}:ro
    - /var/log/accumulo

  gaffer-rest:
    image: gchq/gaffer-rest:${GAFFER_VERSION}-accumulo-${ACCUMULO_VERSION}
    depends_on:
      accumulo-tserver:
        condition: service_healthy
    ports:
    - 8080:8080
    volumes:
    - ./configs/gaffer/application.properties:/gaffer/config/application.properties:ro
    - ./configs/gaffer/data:/gaffer/data:ro
    - ./configs/gaffer/graph:/gaffer/graph:ro
    - ./configs/gaffer/schema:/gaffer/schema:ro
    - ./configs/gaffer/store:/gaffer/store:ro

If you are not familiar with docker or docker compose there are plenty of resources online to get up to speed but essentially, it is a config file that will run up multiple containers with various bits of configuration which is a lot easier than typing out multiple docker commands!

Some key bits you may want to configure in the file are the shared volumes (under the volumes section). The locations in the example file assume you use the project structure from the example setup but if you change any of the locations then they will need updating.

To run up the cluster its as easy as running the following from the root of the project.

docker compose up

The above configuration will start the following containers:

• Zookeeper
• HDFS
  • Datanode
  • Namenode
• Accumulo
  • Monitor
  • GC
  • tServer
  • Master
• Gaffer Rest API

The web UIs for nodes in cluster can then be accessed at the following addresses:

• Access the HDFS NameNode web UI at: http://localhost:9870
• Access the Accumulo Monitor UI at: http://localhost:9995
• Access the Gaffer REST API at: http://localhost:8080/rest/

Environment Variables

As you can probably see the example is using a few environment variables in the docker-compose.yaml file, these set things such as the container versions and a couple of file locations. The use of these variables are recommended as it can make it easier to update container versions and other aspects of the containers.

A basic set of these environment variables are shown below which can be saved in a .env file and sourced before running the containers.

ZOOKEEPER_VERSION="3.7.1"
GAFFER_VERSION="2.0.0"
ACCUMULO_VERSION="2.0.1"
HADOOP_VERSION="3.3.3"
ACCUMULO_CONF_DIR="/etc/accumulo/conf"
HADOOP_CONF_DIR="/etc/hadoop/conf"