July 2017
Beginner to intermediate
378 pages
10h 26m
English
Despite its whimsical name, YARN is the resource allocation component that most Hadoop components rely on. The idea behind YARN was to split resource management away from job scheduling and monitoring. In YARN, these are separate processing units.
As represented in the following diagram, there is a single ResourceManager for the cluster and ApplicationMaster for each job or set of jobs represented by a DAG, which runs on the nodes. The job or DAG of jobs is called an application. There is also a NodeManager for each node in the cluster. It is the job of the NodeManager to monitor the CPU, disk, memory, and network usage of the container on the node where the applications running on it are executed.
Read now
Unlock full access