Once you have loaded your data and tested and tuned your applications, you can add decommission HDFS data-nodes and add them to the MapR cluster.
This is a three-step process:
- Decommissioning nodes on an Apache Hadoop cluster: The Hadoop decommission feature enables you to gracefully remove a set of existing data-nodes from a cluster while it is running, without data loss. For more information, see Hadoop Wiki FAQ.
- Meeting minimum hardware and software requirements: Ensure that every data-node you want to add to the MapR cluster meets the hardware, software, and configuration requirements.
- Adding Nodes to a MapR cluster: You can add those data-nodes to the MapR cluster. For more information, see Adding Nodes to a Cluster.