操作系统为ubuntu10.04,cpu为AMD
1、先到到甲骨文官网下载jdk-7-linux-x64.tar.gz(CPU为其它的注意下载相应的版本),因为hadoop是用java写的,所以要安装java的jdk,我这是把下载的压缩文件解压到目录/usr/lib/jvm/jdk1.7.0,记住这个目录,待会设置hadoop的java环境变量的路径时会用到此目录。
2、再到hadoop官网下载hadoop-0.20.2.tar.gz,解压到某一个目录,我这是 /usr/hadoop-0.20.2
3、修改hadoop的java环境变量的路径,打开/usr/hadoop-0.20.2/conf/hadoop-env.sh文件,添加一行代码:
export JAVA_HOME=/usr/lib/jvm/jdk1.7.0
到此hadoop默认配置完成,但是这个默认的模式是单机非分布式的。
特别注意关于java的安装,参考:http://www.webupd8.org/2011/09/how-to-install-oracle-java-7-jdk-in.html
cd '/home/lxj/Labs' #个人把下载到的文件放在/home/lxj/Labs文件夹里
sudo mkdir -p /usr/lib/jvm/ #just in case (这里新建一个 jvm 文件夹,这步可省略不做)
sudo mv jdk1.7.0/ /usr/lib/jvm/ #生成:/usr/lib/jvm/jdk1.7.0
3. Install Update Java package created by Bruce Ingalls (packages available for Ubuntu 11.10, 11.04, 10.10 and 10.04):
sudo add-apt-repository ppa:nilarimogard/webupd8
sudo apt-get update
sudo apt-get install update-java
If you don't want to add the WebUpd8 PPA, you can also manually download the script.The script may work with other Linux distributions, but it has only been tested on Ubuntu.
4. Now run the following command in a terminal to install Oracle Java JDK:
sudo update-java
The script will ask you to select the Java version - select "jdk1.7.0" for Oracle Java JDK 7, click OK and wait for it to finish:
After a few minutes, Oracle Java JDK should be successfully installed on your Ubuntu machine. You can check out the version by running these commands in a terminal:java -version javac -version
The version should be "1.7.0".