mattjtodd / docker-spark

Apache Spark docker image

Spark docker

Docker images to:

• Setup a standalone Apache Spark cluster running one Spark Master and multiple Spark workers
• Build Spark applications in Java, Scala or Python to run on a Spark cluster

Currently supported versions:

• Spark 2.4.0 for Hadoop 2.8 with OpenJDK 8 and Scala 2.12
• Spark 2.4.0 for Hadoop 2.7+ with OpenJDK 8
• Spark 2.3.2 for Hadoop 2.7+ with OpenJDK 8
• Spark 2.3.1 for Hadoop 2.7+ with OpenJDK 8
• Spark 2.3.1 for Hadoop 2.8 with OpenJDK 8
• Spark 2.3.0 for Hadoop 2.7+ with OpenJDK 8
• Spark 2.2.2 for Hadoop 2.7+ with OpenJDK 8
• Spark 2.2.1 for Hadoop 2.7+ with OpenJDK 8
• Spark 2.2.0 for Hadoop 2.7+ with OpenJDK 8
• Spark 2.1.3 for Hadoop 2.7+ with OpenJDK 8
• Spark 2.1.2 for Hadoop 2.7+ with OpenJDK 8
• Spark 2.1.1 for Hadoop 2.7+ with OpenJDK 8
• Spark 2.1.0 for Hadoop 2.7+ with OpenJDK 8
• Spark 2.0.2 for Hadoop 2.7+ with OpenJDK 8
• Spark 2.0.1 for Hadoop 2.7+ with OpenJDK 8
• Spark 2.0.0 for Hadoop 2.7+ with Hive support and OpenJDK 8
• Spark 2.0.0 for Hadoop 2.7+ with Hive support and OpenJDK 7
• Spark 1.6.2 for Hadoop 2.6 and later
• Spark 1.5.1 for Hadoop 2.6 and later

Using Docker Compose

Add the following services to your docker-compose.yml to integrate a Spark master and Spark worker in your BDE pipeline:

spark-master:
  image: bde2020/spark-master:2.4.0-hadoop2.7
  container_name: spark-master
  ports:
    - "8080:8080"
    - "7077:7077"
  environment:
    - INIT_DAEMON_STEP=setup_spark
    - "constraint:node==<yourmasternode>"
spark-worker-1:
  image: bde2020/spark-worker:2.4.0-hadoop2.7
  container_name: spark-worker-1
  depends_on:
    - spark-master
  ports:
    - "8081:8081"
  environment:
    - "SPARK_MASTER=spark://spark-master:7077"
    - "constraint:node==<yourworkernode>"
spark-worker-2:
  image: bde2020/spark-worker:2.4.0-hadoop2.7
  container_name: spark-worker-2
  depends_on:
    - spark-master
  ports:
    - "8081:8081"
  environment:
    - "SPARK_MASTER=spark://spark-master:7077"
    - "constraint:node==<yourworkernode>"  

Make sure to fill in the INIT_DAEMON_STEP as configured in your pipeline.

Running Docker containers without the init daemon

Spark Master

To start a Spark master:

docker run --name spark-master -h spark-master -e ENABLE_INIT_DAEMON=false -d bde2020/spark-master:2.4.0-hadoop2.7

Spark Worker

To start a Spark worker:

docker run --name spark-worker-1 --link spark-master:spark-master -e ENABLE_INIT_DAEMON=false -d bde2020/spark-worker:2.4.0-hadoop2.7

Launch a Spark application

Building and running your Spark application on top of the Spark cluster is as simple as extending a template Docker image. Check the template's README for further documentation.

• Java template
• Python template
• Scala template