Author: Chuck Lam
Publisher:Manning, 2010
Pages: 325
ISBN: 978-1935182191
Aimed at:
Rating: 5
Pros: Well paced and comprehensive coverage
Cons: No obvious ones
Reviewed by: Mike James

Hadoop isn't a difficult beast to get to grips with but it can seem complicated because there are so many different things that have to be mastered to get it all up and running. This particular book does a good job of taking it slowly enough for you to see how it works and then enabling you to take it futher. 

Chapter 1 is a nice and steady introduction to Hadoop and the whole MapReduce idea. It does this by working through a simple example of word counting and showing how you would split it up into tasks in such a way that MapReduce seems a natural approach. By the end of the chapter you have created and tried out your first Hadoop program. This leaves you wondering what the rest of the book could be about?

Chapter 2 is called Starting Hadoop and this tells you in more detail how to set up, configure and monitor Hadoop. Chapter 3 continues with a look at the components of Hadoop - the Hadoop filing system and  the MapReduce framework.

Part 2 of the book is about Hadoop in action and it goes into more detail and practical considerations of the basic ideas you learned about in the first part. Chapters 4 and 5 takes us deeper into the art of writing MapReduce programs. Chapter 6 looks at the problems and best practices of Hadoop programming - including debugging, logging and performance tuning. Chapter 7 takes the form of a small cookbook with five very common recipes described in some detail. 

Chapter 8 returns to the task of managing Hadoop, but this time considering how to run a production system i.e. what happens when you have lots of machines and a multiple jobs  to manage.  Chapter 9 looks at the alternative approach - the cloud. More sepcificlly it looks at how to use Amazon's AWS system to setup and run a Hadoop system that might cost you a lot of money if you actually bought the hardware.

The final chapters are a collection of untidy topics - programming with Pig as a way of going high level with Hadoop and using Hive for data management.  The book closes with some case studies.

A great book. If you want to get started and more with Hadoop buy, a copy.