hadoop之伪分布搭建环境

1. 修改主机名字  临时修改(session): sudo hostname orcll

/etc/sysconfig/network -> HOSTNAME=hadoop0 #localhost.localdomain    shutdown -r now 2.将hostname与ip绑定

添加/etc/hosts文件 ping hadoop0

3.关闭防火墙

1 service iptables stop
2 chkconfig iptables off
3 --  service iptables status  chkconfig [--list]|grep iptables  

4.免密码登陆

1 ssh-keygen -t rsa
2          id_rsa id_rsa.pub
3 cp id_rsa.pub authorized_keys
4
5 ----  service sshd start  chkconfig sshd on

问题：出现ping baidu.com     connect: Network is unreachable

解决：

1>linux主动添加网关(GATEWAY):   route add default gw 192.168.191.100 dev eth0

2>/etc/sysconfig/network-scriptings/ifcfg-eth0   设置GATEWAY=192.168.191.1

命令：

重启network: service network restart

查看路游表:  route

如果没有安装ssh：    #yum install ssh(先可以访问外网)   yum -y install openssh-clients   yum -y install openssh-server

ssh localhost

5.安转jdk

tar -vxf /opt/hadoop/jdk-8u25-linux-x64.gz.jar  编辑/etc/profile文件，在文件嘴上放添加   export JAVA_HOME=/opt/hadoop/jdk1.8.0_25   export PATH=.:$JAVA_HOME/bin:$PATH  source /etc/profile  java -version

6.安装hadoop

tar -zxvf hadoop-1.1.2.tar.gz  编辑/etc/profile文件  export HADOOP_HOME=/opt/hadoop/hadoop-1.1.2

source /etc/profile

修改hadoop配置文件

hadoop-env.sh

1 设置jdk环境路径，第九行

core.site.xml

1 <property>
2   <name>fs.default.name</name>
3   <value>hdfs://hadoop0:9000</value>
4   <description>change your own hostname</description>
5 </property>
6 <property>
7   <name>hadoop.tmp.dir</name>
8   <value>/usr/local/hadoop-1.1.2/tmp</value>
9 </property>

hdfs-site.xml

1 <property>
2   <name>dfs.replication</name>
3   <value>1</value>
4 </property>
5 <property>
6   <name>dfs.permissions</name>
7   <value>false</value>
8 </property>

mapred-site.xml

1 <property>
2   <name>mapred.job.tracker</name>
3   <value>hadoop0:9001</value>
4   <description>change your own hostname</description>
5 </property>

格式化：hadoop namenode -format 启动：start-all.sh

验证：jps

http://hadoop0:50070/dfshealth.jsp http://hadoop0:50030/jobtracker.jsp