Workload Automation for Hadoop integration is priced by 'Node'. What is a 'node' in this context? Is it a physical/virtual server running an Hadoop component?
A simple Hadoop setup will consist of master and worker nodes. The master node maintains various process such as job and data tracker, name and data node information. The worker aka slave nodes maintain data and may also be compute-only node.
So, in short nodes are hosts in a cluster with various process and perform several activities for the application.