The Definitive Guide to Hadoop
Chapter 4: The Tale of yarn
Using Hadoop 2 exclusively, author Tom White presents a new chapter on YARN and several Hadoop-related projects such as Parquet, Flume, Crunch, and Spark. This chapter covers YARN in depth and will drill further into topics such as:
- The anatomy of a YARN application run
- How to build a YARN application
- Understanding the difference between YARN and MapReduce 1
- Scheduler options in YARN and the capabilities and limitations of each
- Understanding how resources are shared between queues
With YARN, you can run several different applications on a single cluster - and that's great. But with so many simultaneous workloads competing for resources, hitting a performance wall is inevitable. That's where Pepperdata comes in: we let you blast through the performance wall by dynamically shaping the workloads on your cluster to take advantage of previously unused resources.
Enjoy this free chapter on us - and for more information on Pepperdata, click here.