Hadoop学习四:hadoop分布式环境搭建

Hadoop学习四:hadoop分布式环境搭建

标签（空格分隔）： hadoop

---

一，环境准备

1，删除/opt/app/hadoop-2.5.0/share/doc目录（该目录很占空间）

rm -rf doc/

2，关闭虚拟机

[root@hadoop001 hadoop-2.5.0]# halt

3，克隆虚拟机（hadoop002 hadoop003）

4，IP和主机名的关系
hadoop001.com.cn—-192.168.44.100
hadoop002.com.cn—-192.168.44.102
hadoop003.com.cn—-192.168.44.103

二，分布式环境搭建（针对克隆）

1，修改主机名称

[root@hadoop001 Desktop]# vi /etc/sysconfig/network

2，修改mac地址

[root@hadoop001 Desktop]# vi /etc/udev/rules.d/70-persistent-net.rules 

修改后的文件如下：

3，修改IP地址

[root@hadoop001 Desktop]# setup

4, 修改网卡文件配置

[root@hadoop001 Desktop]# vi /etc/sysconfig/network-scripts/ifcfg-eth0

删除UUID和HWADDR后如下：

5，重启机器生效

三，角色分配

namenode
secondarynamenode
datanode
resourcemanager
nodemanager
historyserver

hadoop001.com.cn (namenode datanode nodemanage)
hadoop002.com.cn (datanode nodemanage resourcemanage)
hadoop003.com.cn (datanode nodemanage secondarynamenode historyserver)

四，安装配置

1，修改本地hosts文件

vi /etc/hosts

192.168.44.100 hadoop001.com.cn hadoop001
192.168.44.102 hadoop002.com.cn hadoop002
192.168.44.103 hadoop003.com.cn hadoop003

2,配置无秘钥登陆
hadoop001—>hadoop001,hadoop002,hadoop003
hadoop002—>hadoop001,hadoop002,hadoop003

ssh-keygen -t rsa   //生成秘钥
cd 
cd .ssh/
ssh-copy-id hadoop001.com.cn   //将秘钥拷贝到其他服务器上

3,配置NTP服务器—用于保持各个服务器之间的时间同步
hadoop001.com.cn

 [root@hadoop001 .ssh]# rpm -qa|grep ntp

 vi /etc/ntp.conf

 vi /etc/sysconfig/ntpd 

添加 SYNC_HWCLOCK=yes

修改系统的时间和当前时间一致

 ntpdate 0.centos.pool.ntp.org

启动服务

service ntpd start
chkconfig ntpd on   //开机自动启动

配置其他服务器和hadoop001.com.cn的时间一致

crontab -e
---------
*/10 * * * * /usr/sbin/ntpdate hadoop001.com.cn
---------
service crond restart

4,配置namenode和secondarynamenode
在hadoop001.com.cn服务器上修改hdfs-site.xml文件

<configuration>
    <property>
        <name>dfs.replication</name>
        <value>1</value>
    </property>
    <property>
        <name>dfs.namenode.http-address</name>
        <value>hadoop001.com.cn:50070</value>
    </property>
    <property>
        <name>dfs.namenode.secondary.http-address</name>
        <value>hadoop003.com.cn:50090</value>
    </property>
</configuration>

5,配置datanode
在hadoop001.com.cn服务器上修改slaves文件

hadoop001.com.cn
hadoop002.com.cn
hadoop003.com.cn

6,配置resourcemanager
修改yarn-site.xml文件

<configuration>
    <property>
        <name>yarn.nodemanager.aux-services</name>
        <value>mapreduce_shuffle</value>
    </property>

    <property>
        <name>yarn.log-aggregation-enable</name>
        <value>true</value>
    </property>
    <property>
        <name>yarn.log-aggregation.retain-seconds</name>
        <value>604800</value>
    </property>
    <property>
        <name>yarn.resourcemanager.hostname</name>
        <value>hadoop002.com.cn</value>
    </property>
</configuration>

7,配置historyserver
修改mapred-site.xml文件

<configuration>
    <property>
        <name>mapreduce.framework.name</name>
        <value>yarn</value>
    </property>
    <property>
        <name>mapreduce.jobhistory.webapp.address</name>
        <value>hadoop003.com.cn:19888</value>
    </property>
</configuration>

8,将hadoop001.com.cn服务器上的hadoop安装包复制到hadoop002.com.cn和hadoop003.com.cn上面

[hadoop001@hadoop001 app]$ scp -r hadoop-2.5.0/ hadoop002.com.cn:/opt/app/

[hadoop001@hadoop001 app]$ scp -r hadoop-2.5.0/ hadoop003.com.cn:/opt/app/

9,启动集群

bin/hdfs namenode -format
sbin/start-all.sh

10，查看集群
http://hadoop001.com.cn:50070/dfshealth.html#tab-overview

bin/hdfs dfs -put etc/hadoop/core-site.xml /

bin/hdfs dfs -ls /