Author: Tom White
Publisher: O'Reilly, 2009
Pages: 524
ISBN: 978-0596521974
Aimed at: Developers who already have some familiarity with Hadoop
Rating: 4
Pros: Well written, in depth coverage of practical issues 
Cons: Lacks an overview and unsuitable for beginners to Hadoop
Reviewed by: Mike James

Hadoop is an open source system that allows you to use a cluster of machines to solve a single problem using the MapReduce algorithm. In practice the machines can be off the shelf and nothing special and be linked together using nothing but a standard network. You can even consider using clusters of virtual machines and cloud-based implementations but to make it all work you have to understand the MapReduce algorithm and Hadoop in particular.

This particular book has an elephant on the cover and this is appropriate because the name of the package comes from the name given to a stuffed elephant by one of the project's leading lights. However the book's presentation makes it difficult for the beginner to see the whole of the Hadoop elephant because of the way it spoons it out to you in little chunks.

It starts off with a history of Hadoop and the MapReduce algorithm but without really telling you what either are. Next we move on to implementing a simple MapReduce example and the various parts of Hadoop that are involved - the distributed file system in particular. However even by the end of Chapter Four you still don't have a clear overview of the idea.

If you already know about MapReduce and have some idea of what Haddop is all about then you might not notice the lack of an overview.

The later chapters deal with "higher" level concerns such as configuration, testing, debugging and setting up Hadoop. We also have a description of the inner workings of the algorithm and how Hadoop handles things.

The final part of the book deals with the frameworks built on Hadoop such as Pig, HBase and Zookeeper finishing with some case studies.

All of the coverage is good - well written and dealing with the sort of details that you want to know about - and if you know what the Hadoop elephant looks like this is a useful book. If you are a complete beginner and are struggling even to figure out what the MapReduce algorithm is then it is not a good place to start as it simply never provides a complete overview and even deals with installing Hadoop at the end of the book.

This is not a presentation optimised for the beginner.