YARN

Apache YARN is responsible for managing resources and jobs on a Hadoop Cluster. It includes a global ResourceManager (RM) and per-application ApplicationMaster (AM).

The ResourceManager and the NodeManager form the data-computation framework.

The ResourceManager carves out resources among all the applications in the system. The NodeManager is the per-machine framework agent responsible for containers, monitoring their resource usage (cpu, memory, disk, network) and reporting that back to the ResourceManager/Scheduler.

The ApplicationMaster negotiates resources from the ResourceManager and works with the NodeManager(s) to execute and monitor the tasks.