当前的数据密集型计算需要处理PB级数

据集和GB级数据流，面临着大规模数据管理、复

杂计算环境管理、可扩展计算平台等方面的难题．．

Hadoop是一种易扩展的分布式计算架构，能将廉

价PC节点联合起来提供大型计算服务一其HDFS

提供大规模存储管理，其Map Reduce并行框架为

用户提供容易使用的并行编程模式。本文研究

Hadoop架构并探讨在数据密集型计算中的应用-Current data-intensive computing needs of the PB-level data set, and GB data stream, facing the large-scale data management, complex computing environment management, scalable computing platforms and other aspects of the problem. . Hadoop is a distributed computing architecture for easy expansion, able to inexpensive PC nodes together to provide a large-scale computing services to its HDFS provides large-scale storage management, Map Reduce parallel framework to provide users with easy-to-use parallel programming mode. The Hadoop framework of this study and to explore in the data-intensive computing applications. .