1 tar -zxvf hadoop-1.1.2.tar.gz

2 在hadoop／conf目录

(1) 编辑 hadoop-env.sh

export JAVA_HOME=/usr/java/jdk1.7.0_25

　　然后运行hadoop-env.sh

(2) 编辑 core-site.xml

    <configuration>

　　<property>
     <name>fs.default.name</name>

     <value>hdfs://yjdabc:9000</value>

   </property>

　　<property>

　　　　<name>hadoop.tmp.dir</name>

　　　　<value>/home/yjdabc/hadooptmp</value>

　　</property>

</configuration>

(3)配置hdfs-site.xml

 <configuration>

  <property>

  <name>dfs.name.dir</name>

  <value>/home/yjdabc/hadoop-1.1.2/name</value>

 </property>

 <property>

  <name>dfs.data.dir</name>

  <value>/home/yjdabc/hadoop-1.1.2/data</value>

 </property>

 <property>

  <name>dfs.replication</name>

  <value>1</value>

 </property>

</configuration>
(4)配置mapred-site.xml

<configuration>

  <property>

   <name>mapred.job.tracker</name>

   <value>yjdabc:9001</value>

 </property>
</configuration>

(5)配置masters文件和slaves文件
    将里面内容改为：yjdabc
(6)vi /etc/hosts

127.0.0.1    yjdabc yjdabc.centOSDomain localhost.localdomain localhost

 ::1         yjdabc yjdabc.centOSDomain localhost6.localdomain6 localhost6
yjdabc master
yjdabc slave

3 验证Hadoop(修改配置项后，一定要运行此命令重新配置namenode)
 在hadoop／bin目录下
 ./hadoop namenode -format

4 启动Hadoop
 在hadoop／bin目录下
 ./start-all.sh

5 查看JPS查看hadoop是否完全启动

 [root@localhost bin]# jps

 15219 JobTracker

 15156 SecondaryNameNode

 15495 Jps

 15326 TaskTracker

 15044 DataNode

 14959 NameNode

6 查看集群状态
./hadoop dfsadmin -report