Real-World Hadoop

Chapter 5. Prototypical Hadoop Use Cases

In this chapter, we present a collection of some of the most appealing use cases when considering what Hadoop might do for you. These are prototypical use cases—isolated goals and the Hadoop-based solutions that address them. They are not theoretical. They come from observations of some of the most common and most rewarding ways in which Hadoop is being used, including showing its value in large-scale production systems. We identify each use case by the major intent of its design, to make it easier for you to extrapolate to your own needs. Then, in Chapter 6, we will tell some real-world stories about how MapR customers have combined these use cases to put Hadoop to work.

Data Warehouse Optimization

One of the most straightforward ways to use Hadoop to immediate advantage is to employ it for optimization of your use of a costly data warehouse. The goal of data warehouse (DW) optimization is to make the best use of your data warehouse (or relational database) resources in order to lower costs and keep your data warehouse working efficiently as your data scales up. One way to do this that offers a big payback is to move early ETL processing and staging tables off of the data warehouse onto a Hadoop cluster for processing. This approach is advantageous because these ETL steps often consume the majority of the processing power of the data warehouse, but they only constitute a much smaller fraction of the total lines of code. Moreover, the staging ...