In this chapter, we started with a brief history of Hadoop releases. Next, we covered the basics of Hadoop 1.x and MRv1. We then looked at the core differences between MRv1 and MRv2 and how YARN fits into a Hadoop environment. We also saw how the JobTracker's responsibilities were broken down in Hadoop 2.x.

We also talked about the old and new MapReduce APIs, their origin, differences, and support in YARN. Finally, we concluded the chapter with some practical examples using the old and new MapReduce APIs.

In the next chapter, you will learn about the administration part of YARN.