The previous chapter introduced a gamut of Giraph APIs and showed how you can use them for various graph-processing applications. This chapter pops the hood on Giraph implementation and shows you what transpires behind the scenes when a graph-processing application is executed and APIs are called.
For the rest of this chapter, you use the example Giraph applications introduced in Chapter 5. Assume that the size of the graph on which the application operates is measured in billions of lines of input. With this assumption in place, the chapter traces every step of the application run on a Hadoop cluster. You start ...