安装 jdk
yum install java-1.7.-openjdk*
检查安装：java -version
创建Hadoop用户，设置Hadoop用户使之可以免密码ssh到localhost
su - hadoop
ssh-keygen -t dsa -P '' -f ~/.ssh/id_dsa
cat ~/.ssh/id_dsa.pub>> ~/.ssh/authorized_keys cd /home/hadoop/.ssh
chmod authorized_keys
注意这里的权限问题，保证.ssh目录权限为700，authorized_keys为600
验证：
[hadoop@localhost .ssh]$ ssh localhost
Last login: Sun Nov ::
解压hadoop，安装在/opt/hadoop
tar -xzvf hadoop-2.6..tar.gz
mv -i /home/erik/hadoop-2.6. /opt/hadoop
chown -R hadoop /opt/hadoop
要修改的文件有hadoop-env.sh、core-site.xml 、 hdfs-site.xml 、 yarn-site.xml 、mapred-site.xml几个文件。
1 cd /usr/opt/hadoop/etc/hadoop
设置hadoop-env.sh中的java环境变量,改成这样JAVA_HOME好像没效
export JAVA_HOME= {你的java环境变量}
core-site.xml
<configuration>
<property>
<name>hadoop.tmp.dir</name>
<value>/opt/hadoop/tmp</value>
</property>
<property>
<name>fs.default.name</name>
<value>localhost:</value>
</property>
</configuration>
hdfs.xml
<configuration>
<property>
<name>dfs.replication</name>
<value></value>
</property>
<property>
<name>dfs.namenode.name.dir</name>
<value>/opt/hadoop/dfs/name</value>
</property>
<property>
<name>dfs.datanode.data.dir</name>
<value>/opt/hadoop/dfs/data</value>
</property>
<property>
<name>dfs.permissions</name>
<value>false</value>
</property>
</configuration>
yarn-site.xml
<configuration>
<property>
<name>mapreduce.framework.name</name>
<value>yarn</value>
</property> <property>
<name>yarn.nodemanager.aux-services</name>
<value>mapreduce_shuffle</value>
</property>
</configuration>
mapred-site.xml
<configuration>
<property>
<name>mapred.job.tracker</name>
<value>localhost:</value>
</property>
</configuration>
配置环境变量，修改/etc/profile, 写在最后面即可。配置完要重启！！！
export JAVA_HOME=/usr/lib/jvm/java-1.7.-openjdk-1.7.0.95.x86_64
export JRE_HOME=$JAVA_HOME/jre
export CLASSPATH=.:$JAVA_HOME/lib/dt.jar:$JAVA_HOME/lib/tools.jar:$JAVA_HOME/bin
export HADOOP_INSTALL=/opt/hadoop
export PATH=${HADOOP_INSTALL}/bin:${HADOOP_INSTALL}/sbin${PATH}
export HADOOP_MAPRED_HOME=${HADOOP_INSTALL}
export HADOOP_COMMON_HOME=${HADOOP_INSTALL}
export HADOOP_HDFS_HOME=${HADOOP_INSTALL}
export YARN_HOME=${HADOOP_INSTALLL}
export HADOOP_COMMON_LIB_NATIVE_DIR=${HADOOP_INSTALL}/lib/natvie
export HADOOP_OPTS="-Djava.library.path=${HADOOP_INSTALL}/lib:${HADOOP_INSTALL}/lib/native"
之后就是见证奇迹的时候了，
cd /opt/hadoop/
格式化hdfs
bin/hdfs namenode -format
启动hdfs
sbin/start-dfs.sh
sbin/start-yarn.sh
理论上会见到
Starting namenodes on [localhost]
localhost: starting namenode, logging to /usr/opt/hadoop-2.6./logs/hadoop-hadoop-namenode-.out
localhost: starting datanode, logging to /usr/opt/hadoop-2.6./logs/hadoop-hadoop-datanode-.out
Starting secondary namenodes [0.0.0.0]
0.0.0.0: starting secondarynamenode, logging to /usr/opt/hadoop-2.6./logs/hadoop-hadoop-secondarynamenode-.out
输入网址127.0.0.1:50070就可以看见hadoop的网页了，这就说明成功了。
参考：
http://www.centoscn.com/hadoop/2015/0118/4525.html
http://blog.csdn.net/yinan9/article/details/16805275
http://www.aboutyun.com/thread-10554-1-1.html