Hadoop in Action

This book is the Chinese translation of Hadoop in Action, aimed at programmers, architects, and project managers who need to process massive offline data. It explains the Hadoop framework and MapReduce programming progressively, from the basics to advanced topics. The book is divided into three parts. Part 1 introduces Hadoop's origins, basic concepts, and how to obtain and install a cluster, starting with a simple word-count example to give readers their first impression of the MapReduce computing model. Part 2 delves into the core components of the Hadoop framework, explaining the working principles of the HDFS distributed file system, the execution flow of MapReduce jobs, task scheduling and fault tolerance, and common MapReduce programming patterns and best practices. Part 3 expands the view beyond Hadoop to the wider ecosystem, introducing the Streaming API and higher-level tools such as Pig and Hive that make data processing easier, and discussing how Hadoop is applied in real data analysis tasks. The book is example-driven, with multiple sets of runnable code and small exercises, written in concise, engineering-oriented language. It is suitable as the first systematic book for Hadoop beginners and engineering practice. Because it was written during the Hadoop 1.x era, it does not cover YARN or later ecosystem developments; readers should consult their actual framework versions for comparison.