Installing Hadoop on Linux

Prerequisites: Hadoop is written in Java, so install Java first. Installing the JDK on Ubuntu see: http://blog.csdn.net/microfhu/article/details/7667393
The Hadoop version I downloaded is 2.4.1, which requires at least JDK 6 to be installed. Linux is the only supported production environment, unix,windows or Mac OS can be used as a development environment. Installing Hadoop on Windows requires Cygwin, and includes openssh in Cygwin.
Install Hadoop on Linux to hadoop.apache.org download the latest stable version of hadoop,hadoop-2.4.1.tar.gz to extract it to a directory, for example:/home/norris/soft/ hadoop-2.4.1 decompression command: TAR-XVF./hadoop-2.4.1.tar.gz Configure Environment variables: Open/home/norris/.profile file: VI ~/.profile I'm using Ubuntu. 14.04.1, this file for other Linux systems may be called. Bash_profile or. Bash_login at the end of the add: Export hadoop_install=/home/norris/soft/ Hadoop-2.4.1export Path=${path}:${hadoop_install}/bin:${hadoop_install}/sbin saves the file and executes it into effect: the source ~/. Profile at this point you should have Java and hadoop_install in your environment variables
Execution: Hadoop version, which appears with the release information, indicates a successful installation. The next step is to configure the operating mode and hdfs,mapreduce for Hadoop (Yarn)