Apache Hadoop Container

Build An Apache Hadoop Container

$] docker build -t voyant/apache-hadoop .

Deploying A Hadoop Container

The following environment variables are used to initialize an Apache Hadoop Container.

• HADOOP_CLUSTER_NAME This variable contains the Apache Hadoop cluster name and is required when $HADOOP_NAME_NODE_INIT is set to true. (Optional)
• HADOOP_NAME_NODE_INIT This variable when set to true will bootstrap the Apache Hadoop cluster. (Optional)
• HADOOP_NODE_TYPE This variable contains the specialization of a container. Possible options are datanode, historyserver, namenode, nodemanager, resourcemanager and webappproxy.
• SSHD_PORT This variable is only valid for node types datanode and nodemanager. It containes the port number the container's SSHd should listen on.

To run a name node container just run the following command.

$] sudo docker run -d --name namenode -e HADOOP_NODE_TYPE=namenode -v /etc/hadoop:/etc/hadoop voyant/apache-hadoop

Note: If the host folder /etc/hadoop is empty the container will write a set of default config files and exit. Once the files are updated restart the container and the selected specialization will run with the updated configuration.

Note: The container expects the logs to be stored in the /var/log/hadoop folder. Keep that in mind when updating the configuration files.

For more information on how to deploy an Apache Hadoop cluster please visit the project site.