首先 编辑 hosts文件
vi /etc/hosts
根据你的集群架构进行配置
192.168.101.11 bigdata11
192.168.101.12 bigdata12
192.168.101.13 bigdata13
192.168.101.14 bigdata14
192.168.101.15 bigdata15
参照之前文章在bigdata15上安装hadoop
参见之前文章 将每台机器的公钥部署在其他机器上
需要在每一台机器上执行如下命令
ssh-copy-id -i /root/.ssh/id_rsa.pub root@bigdata11
ssh-copy-id -i /root/.ssh/id_rsa.pub root@bigdata12
ssh-copy-id -i /root/.ssh/id_rsa.pub root@bigdata13
ssh-copy-id -i /root/.ssh/id_rsa.pub root@bigdata14
ssh-copy-id -i /root/.ssh/id_rsa.pub root@bigdata15
进入bigdata15 hadoop配置目录
cd /etc/hadoop/etc/hadoop/
vi hdfs-site.xml
<property>
<name>dfs.replication</name>
<value>3</value>
</property>
将每台机器的冗余设为3
core-site.xml 与伪分布一致 如下(NameNode地址)
<configuration>
<property>
<name>fs.defaultFS</name>
<value>hdfs://bigdata15:9000</value>
</property>
<property>
<name>hadoop.tmp.dir</name>
<value>/hdfs/tmp</value>
</property>
</configuration>
注意在每台机器上创建该目录
mkdir /hdfs
mkdir /hdfs/tmp
mapred-site.xml 与伪分布一致
创建该文件
cp mapred-site.xml.template mapred-site.xml
编辑内容如下
<configuration>
<property>
<name>mapreduce.framework.name</name>
<value>yarn</value>
</property>
</configuration>
yarn-site.mxl 需更改yarn.resourcemanager.hostname的值为主节点地址
<configuration>
<property>
<name>yarn.resourcemanager.hostname</name>
<value>bigdata15</value>
</property>
<property>
<name>yarn.nodemanager.aux-services</name>
<value>mapreduce_shuffle</value>
</property>
</configuration>
编辑slaves
vi slaves
bigdata11
bigdata12
bigdata13
bigdata14
格式化
hdfs namenode -format
复制配置文件至其他主机
scp -r /etc/hadoop/ root@bigdata11:/etc/
scp -r /etc/hadoop/ root@bigdata12:/etc/
scp -r /etc/hadoop/ root@bigdata13:/etc/
scp -r /etc/hadoop/ root@bigdata14:/etc/
注意更改防火墙,也可以使用远程拷贝
scp -r /etc/sysconfig/iptables bigdata11:/etc/sysconfig/iptables
scp -r /etc/sysconfig/iptables bigdata12:/etc/sysconfig/iptables
scp -r /etc/sysconfig/iptables bigdata13:/etc/sysconfig/iptables
scp -r /etc/sysconfig/iptables bigdata14:/etc/sysconfig/iptables
搞定启动服务
start-all.sh
查看集群状态
hadoop dfsadmin -report