Hadoop实现了一个分布式文件系统（Hadoop Distributed File System），简称HDFS。HDFS有着高容错性的特点，并且设计用来部署在低廉的（low-cost）硬件上。而且它提供高传输率（high throughput）来访问应用程序的数据，适合那些有着超大数据集（large data set）的应用程序。HDFS放宽了（relax）POSIX的要求（requirements）这样可以流的形式访问（streaming access）文件系统中的数据。-Hadoop implements a distributed file system (Hadoop Distributed File System), referred to HDFS. HDFS has a high fault tolerance features, and designed to be deployed in low (low-cost) hardware. And it offers high transfer rate (high throughput) to access the application data, for those with very large data sets (large data set) applications. HDFS relax (relax) POSIX requirements (requirements) that can flow in the form of access (streaming access) file system data.