hadoop伪分布模式的安装

                                                                                  hadoop 伪分布模式的安装

1.修改配置文件

        1）修改 core-site.xml

            <configuration>
            <property>
            <name>fs.default.name</name>
            <value>hdfs://localhost:9000</value>
            </property>
            </configuration>

        2）修改hadoop-env.sh

              将 #export JAVA_HOME=/usr/lib /java

              改为 export JAVA_HOME= /usr/lib/jdk/jdk1.8.0_20

              /usr/lib/jdk/jdk1.8.0_20 为计算机环境的java安装目录中的bin目录的父级目录

        3）修改mapred-site.xml
           <configuration>
           <property>
           <name>mapred.job.tracker</name>
           <value>localhost:9001</value>
           </property>
           </configuration>
 

       4）修改hdfs-site.xml

           <configuration>
           <property>
           <name>dfs.data.dir</name>
           <value>/usr/hadoop/data</value>
           </property>
           <property>
           <name>dfs.replication</name>
           <value>1</value>
           </property>
           </configuration>

2.生成密钥对

         1)#cd  /root

         2)#ssh-keygen -t rsa

           Generating public/private rsa key pair.

           Enter file in which to save the key (/root/.ssh/id_rsa):

           /root/.ssh/id_ras already exists.

          Overwrite (y/n)? y

          3)#cd ./ssh

          4)#ls 

            id_rsa   id_rsa.pub  authorized_keys

          5)#cp id_rsa.pub authorized_keys

              cp: overwrite 'authorized_keys' ?      y

3.格式化分布式文件系统

     进入 hadoop的解压目录

     $ bin/hadoop namenode -format

 4.启动守护进程

      进入hadoop的解压目录

      $bin/start-all.sh

5.检测守护进程的启动情况

       进入解压hadoop目录

       1）$ ls /usr/lib/jdk/jdk1.8.0_20/bin/

        2)$/usr/lib/jdk/jdk1.8.0_20/bin/jps

 

 

 

这样 一个hadoop伪分布模式就安装好了