Hadoop ha

Hadoop HA部署（防火墙全部关闭）

修改三台虚拟机主机名

[root@localhost ~]# hostnamectl set-hostname master
[root@localhost ~]# bash

[root@localhost ~]# hostnamectl set-hostname slave1
[root@localhost ~]# bash

[root@localhost ~]# hostnamectl set-hostname slave2
[root@localhost ~]# bash

配置三台虚拟机IP与主机名映射

[root@master ~]# vi /etc/hosts

分发hosts文件到slave

[root@master ~]# scp /etc/hosts slave1:/etc

[root@master ~]# scp /etc/hosts slave2:/etc

配置三台虚拟机SSH无密登录

生成密钥对

[root@master ~]# ssh-keygen -t rsa -P ''

[root@slave1 ~]# ssh-keygen -t rsa -P ''

[root@slave2 ~]# ssh-keygen -t rsa -P ''

复制公钥到slave

[root@master ~]# ssh-copy-id master

[root@master ~]# ssh-copy-id slave1

[root@master ~]# ssh-copy-id slave2

测试从master登录到slave1、slave2

[root@master ~]# ssh slave1

[root@slave1 ~]# exit

[root@master ~]# ssh slave2

[root@slave2 ~]# exit

安装配置jdk

检查三台虚拟机是否有openjdk，如若有需要将其卸载掉，没有可进行下一步操作

[root@master ~]# rpm -qa | grep openjdk

卸载命令

rpm -e --nodeps +查询出来的openjdk包

解压缩jdk包到指定目录

[root@master ~]# tar -zxvf /h3cu/jdk-8u144-linux-x64.tar.gz -C /usr/local/src

配置jdk环境变量

[root@master ~]# vi /root/.bash_profile

添加配置：

export JAVA_HOME=/usr/local/src/jdk1.8.0_144
export JRE_HOME=/usr/local/src/jdk1.8.0_144/jre
export CLASSPATH=.:$CLASSPATH:$JAVA_HOME/lib:$JRE_HOME/lib
export PATH=$PATH:$JAVA_HOME/bin:$JRE_HOME/bin

生效环境变量

[root@master ~]# source /root/.bash_profile

复制分发jdk环境变量文件到slave

[root@master ~]# scp -r /root/.bash_profile slave1:/root

[root@master ~]# scp -r /root/.bash_profile slave2:/root

复制分发jdk安装包文件到slave

[root@master ~]# scp -r /usr/local/src/jdk1.8.0_144/ slave1:/usr/local/src

[root@master ~]# scp -r /usr/local/src/jdk1.8.0_144/ slave2:/usr/local/src

生效环境变量

[root@slave1 ~]# source /root/.bash_profile

[root@slave2 ~]# source /root/.bash_profile

查看jdk版本

[root@slave1 ~]# java -version

[root@slave2 ~]# java -version

安装配置zookeeper

解压缩

[root@master ~]# tar -zxvf /h3cu/zookeeper-3.4.8.tar.gz -C /usr/local/src

重命名

[root@master ~]# mv /usr/local/src/zookeeper-3.4.8/ /usr/local/src/zookeeper

配置环境变量

[root@master ~]# vi /root/.bash_profile

添加配置

export ZOOKEEPER_HOME=/usr/local/src/zookeeper
export PATH=$ZOOKEEPER_HOME/bin:$PATH

分发环境变量文件

[root@master ~]# scp -r /root/.bash_profile  slave1:/root

[root@master ~]# scp -r /root/.bash_profile  slave2:/root

生效环境变量文件

[root@master ~]# source /root/.bash_profile

[root@slave1 ~]# source /root/.bash_profile

[root@slave2 ~]# source /root/.bash_profile

修改zoo.cfg文件

切换到zookeeper配置文件目录

[root@master ~]# cd /usr/local/src/zookeeper/conf/

[root@master conf]# cp zoo_sample.cfg  zoo.cfg

[root@master conf]# vi zoo.cfg

修改

dataDir=/usr/local/src/zookeeper/data

增加

dataLogDir=/usr/local/src/zookeeper/logs

server.1=master:2888:3888

server.2=slave1:2888:3888

server.3=slave2:2888:3888

创建myid配置文件

[root@master conf]# mkdir /usr/local/src/zookeeper/data

[root@master conf]# cd /usr/local/src/zookeeper/data

[root@master data]# echo '1' > myid

分发zookeeper文件

[root@master data]# scp -r /usr/local/src/zookeeper/ slave1:/usr/local/src

[root@master data]# scp -r /usr/local/src/zookeeper/ slave2:/usr/local/src

修改slave1、slave2的myid

[root@slave1 ~]# cd /usr/local/src/zookeeper/data/
[root@slave1 data]# echo '2' > myid

[root@slave2 ~]# cd /usr/local/src/zookeeper/data/
[root@slave2 data]# echo '3' > myid

启动zookeeper

[root@master ~]# zkServer.sh start

[root@slave1 data]# zkServer.sh start

[root@slave2 data]# zkServer.sh start

查看slave2状态

[root@slave2 data]# zkServer.sh status

配置hadoop ha

解压缩

[root@master ~]# tar -zxvf /h3cu/hadoop-2.7.1.tar.gz -C /usr/local/src

重命名

[root@master ~]# mv /usr/local/src/hadoop-2.7.1/ /usr/local/src/hadoop

配置环境变量并分发生效

[root@master ~]# vi /root/.bash_profile

添加配置：

export HADOOP_HOME=/usr/local/src/hadoop
export PATH=$HADOOP_HOME/bin:$HADOOP_HOME/sbin:$PATH

分发文件

[root@master ~]# scp -r /root/.bash_profile slave1:/root

[root@master ~]# scp -r /root/.bash_profile slave2:/root

生效环境变量文件

[root@master ~]# source /root/.bash_profile

[root@slave1 ~]# source /root/.bash_profile

[root@slave2 ~]# source /root/.bash_profile

配置hadoop配置文件

切换到配置目录下

[root@master ~]# cd  /usr/local/src/hadoop/etc/hadoop/

[root@master hadoop]# vi hadoop-env.sh

修改jdk路径

export JAVA_HOME=/usr/local/src/jdk1.8.0_144`

[root@master hadoop]# vi core-site.xml

添加配置：

<property>
  <name>fs.defaultFS</name>
   <value>hdfs://mycluster</value>

</property>

<property>
    <name>hadoop.tmp.dir</name>
    <value>file:/usr/local/src/hadoop/tmp</value>
</property>

<property>
    <name>ha.zookeeper.quorum</name>
    <value>master:2181,slave1:2181,slave2:2181</value>
</property>

[root@master hadoop]# vi hdfs-site.xml

添加配置：

<property>
   <name>dfs.qjournal.start-segment.timeout.ms</name>
    <value>60000</value>
</property>

<property>
    <name>dfs.nameservices</name>
    <value>mycluster</value>
</property>

 <property>
    <name>dfs.ha.namenodes.mycluster</name>
    <value>master,slave1</value>
</property>

<property>
    <name>dfs.namenode.rpc-address.mycluster.master</name>
    <value>master:8020</value>
</property>

<property>
    <name>dfs.namenode.rpc-address.mycluster.slave1</name>
    <value>slave1:8020</value>
</property>

<property>
    <name>dfs.namenode.http-address.mycluster.master</name>
    <value>master:50070</value>
</property>

<property>
    <name>dfs.namenode.http-address.mycluster.slave1</name>
    <value>slave1:50070</value>
</property>


<property>
      <name>dfs.namenode.shared.edits.dir</name>
<value>qjournal://master:8485;slave1:8485;slave2:8485/mycluster</value>

</property>

<property>

     <name>dfs.client.failover.proxy.provider.mycluster</name>

     <value>org.apache.hadoop.hdfs.server.namenode.ha.ConfiguredFailoverProxyProvider</value>

</property>

<property>
    <name>dfs.ha.fencing.methods</name>
    <value>
     sshfence
     shell(/bin/true)
    </value>
</property>

<property>
   <name>dfs.permissions.enabled</name>
   <value>false</value>
</property>

<property>
    <name>dfs.support.append</name>
    <value>true</value>
</property>


<property>
    <name>dfs.ha.fencing.ssh.private-key-files</name>
    <value>/root/.ssh/id_rsa</value>
</property>


<property>
    <name>dfs.replication</name>
    <value>2</value>
</property>

<property>
    <name>dfs.namenode.name.dir</name>
    <value>/usr/local/src/hadoop/tmp/hdfs/nn</value>
</property>

<property>
    <name>dfs.datanode.data.dir</name>
    <value>/usr/local/src/hadoop/tmp/hdfs/dn</value>
</property>


<property>
    <name>dfs.journalnode.edits.dir</name>
    <value>/usr/local/src/hadoop/tmp/hdfs/jn</value>
</property>

<property>
    <name>dfs.ha.automatic-failover.enabled</name>
    <value>true</value>
</property>

<property>
    <name>dfs.ha.fencing.ssh.connect-timeout</name>
    <value>30000</value>
</property>
 

[root@master hadoop]# cp mapred-site.xml.template mapred-site.xml

[root@master hadoop]# vi mapred-site.xml

添加配置：

<property>
    <name>mapreduce.framework.name</name>
    <value>yarn</value>
</property>

<property>
    <name>mapreduce.jobhistory.address</name>
    <value>master:10020</value>
</property>

<property>
    <name>mapreduce.jobhistory.webapp.address</name>
    <value>master:19888</value>
</property>

[root@master hadoop]# vi yarn-site.xml

添加配置：

<property>

         <name>yarn.resourcemanager.ha.enabled</name>

             <value>true</value>

        </property>

        <property>

              <name>yarn.resourcemanager.cluster-id</name>

              <value>yrc</value>

        </property>

        <property>

              <name>yarn.resourcemanager.ha.rm-ids</name>

              <value>rm1,rm2</value>

        </property>

        <property>

              <name>yarn.resourcemanager.hostname.rm1</name>

              <value>master</value>

        </property>

        <property>

              <name>yarn.resourcemanager.hostname.rm2</name>

              <value>slave1</value>

        </property>

        <property>

              <name>yarn.resourcemanager.webapp.address.rm1</name>

              <value>master:8088</value>

        </property>

        <property>

              <name>yarn.resourcemanager.webapp.address.rm2</name>

              <value>slave1:8088</value>

        </property>

        <property>

              <name>hadoop.zk.address</name>

              <value>master:2181,slave1:2181,slave2:2181</value>

        </property>

[root@master hadoop]# vi slaves

添加配置：

master

slave1

slave2

namenode、datanode、journalnode等存放数据的公共目录为/usr/local/src/hadoop/tmp；

 在master上执行如下：

[root@master hadoop]#  mkdir -p /usr/local/src/hadoop/tmp/hdfs/nn

[root@master hadoop]#  mkdir -p /usr/local/src/hadoop/tmp/hdfs/dn

[root@master hadoop]#  mkdir -p /usr/local/src/hadoop/tmp/hdfs/jn

[root@master hadoop]#  mkdir -p /usr/local/src/hadoop/tmp/logs

分发文件

[root@master hadoop]# scp -r /usr/local/src/hadoop root@slave1:/usr/local/src/

[root@master hadoop]# scp -r /usr/local/src/hadoop root@slave2:/usr/local/src/

journalNode初始化和启动

[root@master hadoop]#  hadoop-daemons.sh start journalnode

格式化namenode

[root@master hadoop]# hdfs namenode -format

注册ZNode

[root@master hadoop]# hdfs zkfc -formatZK

复制namenode元数据到其它节点(在master节点执行)

[root@master hadoop]# scp -r /usr/local/src/hadoop/tmp/hdfs/nn/* slave1:/usr/local/src/hadoop/tmp/hdfs/nn/

启动hdfs

[root@master hadoop]# start-all.sh

若没有resourcemanager，就自己手动重启一下

yarn-daemon.sh start resourcemanager

在slave1上启动resourcemanager和namenode进程

[root@slave1 hadoop]#  yarn-daemon.sh start resourcemanager

启动 MapReduce任务历史服务器

[root@master hadoop]# yarn-daemon.sh start proxyserver
[root@master hadoop]# mr-jobhistory-daemon.sh start historyserver

查看master和slave1的进程

[root@master hadoop]# jps

 

[root@slave1 hadoop]# jps