hadoopHA搭建

上一篇安装hadoop集群环境中复制一个出来；

1、cd /opt/module

mkdir HA

chmod 777 HA/

cp -r /opt/module/hadoop-2.7.2 /opt/module/HA/

2、在HA目录操作

============core-site.xml==================

<configuration>
<!-- 把两个NameNode）的地址组装成一个集群mycluster -->
    <property>
        <name>fs.defaultFS</name>
            <value>hdfs://mycluster</value>
    </property>

    <!-- 声明journalnode服务本地文件系统存储目录-->
    <property>
        <name>dfs.journalnode.edits.dir</name>
        <value>/opt/module/HA/hadoop-2.7.2/data/jn</value>
    </property>

    <!-- 指定hadoop运行时产生文件的存储目录 -->
    <property>
        <name>hadoop.tmp.dir</name>
        <value>/opt/module/HA/hadoop-2.7.2/data/tmp</value>
    </property>
    <property>
           <name>ipc.client.connect.max.retries</name>
            <value>20</value>
            <description>
      Indicates the number of retries a clientwill make to establisha server connection.
            </description>
      </property>
      <property>
           <name>ipc.client.connect.retry.interval</name>
            <value>5000</value>
            <description>
      Indicates the number of milliseconds aclient will wait for before retrying to establish a server connection.
            </description>
      </property>
    <property>
        <name>ha.zookeeper.quorum</name>
        <value>hadoop108:2181,hadoop109:2181,hadoop110:2181</value>
    </property>
</configuration>

 

============dfs.hostsl==================

hadoop108
hadoop109
hadoop110

 

============slaves==================

hadoop108
hadoop109
hadoop110

 

============hdfs-site.xml==================

<configuration>
    <!-- 完全分布式集群名称 -->
    <property>
        <name>dfs.nameservices</name>
        <value>mycluster</value>
    </property>

    <!-- 集群中NameNode节点都有哪些 -->
    <property>
        <name>dfs.ha.namenodes.mycluster</name>
        <value>nn1,nn2</value>
    </property>

    <!-- nn1的RPC通信地址 -->
    <property>
        <name>dfs.namenode.rpc-address.mycluster.nn1</name>
        <value>hadoop108:8020</value>
    </property>

    <!-- nn2的RPC通信地址 -->
    <property>
        <name>dfs.namenode.rpc-address.mycluster.nn2</name>
        <value>hadoop109:8020</value>
    </property>

    <!-- nn1的http通信地址 -->
    <property>
        <name>dfs.namenode.http-address.mycluster.nn1</name>
        <value>hadoop108:50070</value>
    </property>

    <!-- nn2的http通信地址 -->
    <property>
        <name>dfs.namenode.http-address.mycluster.nn2</name>
        <value>hadoop109:50070</value>
    </property>

    <!-- 指定NameNode元数据在JournalNode上的存放位置 -->
    <property>
        <name>dfs.namenode.shared.edits.dir</name>
        <value>qjournal://hadoop108:8485;hadoop109:8485;hadoop110:8485/mycluster</value>
    </property>

    <!-- 配置隔离机制，即同一时刻只能有一台服务器对外响应 -->
    <property>
        <name>dfs.ha.fencing.methods</name>
        <value>sshfence</value>
    </property>

    <!-- 使用隔离机制时需要ssh无秘钥登录-->
    <property>
        <name>dfs.ha.fencing.ssh.private-key-files</name>
        <value>/home/hadoop/.ssh/id_rsa</value>
    </property>

    <!-- 关闭权限检查-->
    <property>
        <name>dfs.permissions.enable</name>
        <value>false</value>
    </property>

    <!-- 访问代理类：client，mycluster，active配置失败自动切换实现方式-->
    <property>
          <name>dfs.client.failover.proxy.provider.mycluster</name>
    <value>org.apache.hadoop.hdfs.server.namenode.ha.ConfiguredFailoverProxyProvider</value>
    </property>
    <property>
    <name>dfs.ha.automatic-failover.enabled</name>
    <value>true</value>
</property>

</configuration>

============mapred-site.xml==================

<configuration>
<!--指定mr运行在yarn上 -->
<property>
    <name>mapreduce.framework.name</name>
    <value>yarn</value>    
</property>
<!--历史服务器 -->
<property>
<name>mapreduce.jobhistory.address</name>
<value>hadoop108:10020</value>
</property>
<property>
    <name>mapreduce.jobhistory.webapp.address</name>
    <value>hadoop108:19888</value>
</property>
<!--mapper 输出开启 -->
<property>
    <name>mapreduce.map.output.compress</name>
    <value>true</value>
</property>
<!--使用哪一种压缩方式 -->
<property>
    <name>mapreduce.map.output.compress.codec</name>
    <value>org.apache.hadoop.io.compress.DefaultCodec</value>
</property>

<!--reducer 输出开启 -->
<property>
    <name>mapreduce.output.fileoutputformat.compress</name>
    <value>true</value>
</property>
<!--使用哪一种压缩方式 -->
<property>
    <name>mapreduce.output.fileoutputformat.compress.codec</name>
    <value>org.apache.hadoop.io.compress.DefaultCodec</value>
</property>
<!--SequenceFile 输出使用的压缩类型：NONE & BLOCK-->
<property>
    <name>mapreduce.output.fileoutputformat.compress.type</name>
    <value>RECORD</value>
</property>

</configuration>

 

============yarn-site.xml==================

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

    <!--启用resourcemanager ha-->
    <property>
        <name>yarn.resourcemanager.ha.enabled</name>
        <value>true</value>
    </property>
 
    <!--声明两台resourcemanager的地址-->
    <property>
        <name>yarn.resourcemanager.cluster-id</name>
        <value>cluster-yarn1</value>
    </property>

    <property>
        <name>yarn.resourcemanager.ha.rm-ids</name>
        <value>rm1,rm2</value>
    </property>

    <property>
        <name>yarn.resourcemanager.hostname.rm1</name>
        <value>hadoop108</value>
    </property>

    <property>
        <name>yarn.resourcemanager.hostname.rm2</name>
        <value>hadoop109</value>
    </property>
 
    <!--指定zookeeper集群的地址-->
    <property>
        <name>yarn.resourcemanager.zk-address</name>
        <value>hadoop108:2181,hadoop109:2181,hadoop110:2181</value>
    </property>

    <!--启用自动恢复-->
    <property>
        <name>yarn.resourcemanager.recovery.enabled</name>
        <value>true</value>
    </property>
 
    <!--指定resourcemanager的状态信息存储在zookeeper集群-->
    <property>
        <name>yarn.resourcemanager.store.class</name>
    <value>org.apache.hadoop.yarn.server.resourcemanager.recovery.ZKRMStateStore</value>
</property>

</configuration>

=============修改之后同步其他两台=====================

xsync /opt/module/hadoop-2.7.2 /opt/module/HA/

=====================xsync 执行脚本=============================

#!/bin/bash
#1 获取输入参数个数，如果没有参数，直接退出
pcount=$#
if((pcount==0));then
echo no args;
exit;
fi
#2 获取文件名称
p1=$1
fname=`basename $p1`
echo fname=$fname

#3 获取上级目录到绝对路径
pdir=`cd -P $(dirname $p1); pwd`
echo pdir=$pdir

#4获取当前用户名称
user=`whoami`
#5循环
for((host=109;host<=110;host++));do
    #echo $pdir/$fname $user@hadoop$host:$pdir
    echo --------------hadoop$host----------------
    rsync -rvl $pdir/$fname $user@hadoop$host:$pdir
done

=================初始化============================

#!/bin/bash
echo "================     初始化            ==========="

#ssh hadoop108 'cd /opt/module/HA/hadoop-2.7.2 ; rm -rfv data/ logs/ '
#ssh hadoop109 'cd /opt/module/HA/hadoop-2.7.2 ; rm -rfv data/ logs/ '
#ssh hadoop110 'cd /opt/module/HA/hadoop-2.7.2 ; rm -rfv data/ logs/ '
echo "=========在各个JournalNode节点上，输入以下命令启动journalnode服务=========="
ssh hadoop108 '/opt/module/HA/hadoop-2.7.2/sbin/hadoop-daemon.sh start journalnode'
ssh hadoop109 '/opt/module/HA/hadoop-2.7.2/sbin/hadoop-daemon.sh start journalnode'
ssh hadoop110 '/opt/module/HA/hadoop-2.7.2/sbin/hadoop-daemon.sh start journalnode'

echo "============在[nn1]上，对其进行格式化，并启动===================="
ssh hadoop108 '/opt/module/HA/hadoop-2.7.2/bin/hdfs namenode -format'
ssh hadoop108 '/opt/module/HA/hadoop-2.7.2/sbin/hadoop-daemon.sh start namenode'

echo "=====================在[nn2]上，同步nn1的元数据信息=========================="
ssh hadoop109 '/opt/module/HA/hadoop-2.7.2/bin/hdfs namenode -bootstrapStandby'
echo "================启动[nn2]==================================="
ssh hadoop109 '/opt/module/HA/hadoop-2.7.2/sbin/hadoop-daemon.sh start namenode'

echo "======================关闭namenode=============================="
ssh hadoop108 '/opt/module/HA/hadoop-2.7.2/sbin/hadoop-daemon.sh stop namenode'
ssh hadoop109 '/opt/module/HA/hadoop-2.7.2/sbin/hadoop-daemon.sh stop namenode'

 

 

======================关闭脚本===========================

#!/bin/bash
echo "================     开始关闭所有节点服务            ==========="
echo "================     正在关闭YARN                    ==========="
ssh hadoop108 '/opt/module/HA/hadoop-2.7.2/sbin/stop-yarn.sh'
ssh hadoop109 '/opt/module/HA/hadoop-2.7.2/sbin/yarn-daemon.sh stop resourcemanager'

 

#echo "============各个JournalNode节点上，输入以下命令关闭journalnode服务===================="
#ssh hadoop108 '/opt/module/HA/hadoop-2.7.2/sbin/hadoop-daemon.sh stop journalnode'
#ssh hadoop109 '/opt/module/HA/hadoop-2.7.2/sbin/hadoop-daemon.sh stop journalnode'
#ssh hadoop109 '/opt/module/HA/hadoop-2.7.2/sbin/hadoop-daemon.sh stop journalnode'

echo "================     正在关闭Zookeeper               ==========="
for i in hadoop108 hadoop109 hadoop110
do
         ssh $i 'source /etc/profile ;cd /opt/module/zookeeper-3.4.10/bin/ ; ./zkServer.sh stop'
        echo $i 执行完
done

echo "==============关闭DFSZK Failover Controller============="
ssh hadoop108 '/opt/module/HA/hadoop-2.7.2/sbin/hadoop-daemon.sh stop zkfc'
ssh hadoop109 '/opt/module/HA/hadoop-2.7.2/sbin/hadoop-daemon.sh stop zkfc'
echo "==================关闭HDFS==========================="
ssh hadoop108 '/opt/module/HA/hadoop-2.7.2/sbin/stop-dfs.sh'

echo "======================查看服务状态==================="

for i in hadoop108 hadoop109 hadoop110
do
    echo "===========        $i        ==========="
    ssh $i '/opt/module/jdk1.8.0_191/bin/jps'
done

echo "-------------------------架构-------------------------------- ------------"
echo "-----hadoop108------------hadoop109---------------hadoop110- -------------"
echo "-----NameNode-------------NameNode----------------------------------------"
echo "-----JournalNode----------JournalNode-------------JournalNode-------------"
echo "-----DataNode-------------DataNode----------------DataNode----------------"
echo "-----ZK-------------------ZK----------------------ZK----------------------"
echo "-----ResourceManager------ResourceManager---------------------------------"
echo "-----NodeManager----------NodeManager-------------NodeManager-------------"

 

 

=====================群起脚本=========================

#!/bin/bash
echo "================     开始启动所有节点服务            ==========="

#echo "============各个JournalNode节点上，输入以下命令启动journalnode服务===================="
#ssh hadoop108 '/opt/module/HA/hadoop-2.7.2/sbin/hadoop-daemon.sh start journalnode'
#ssh hadoop109 '/opt/module/HA/hadoop-2.7.2/sbin/hadoop-daemon.sh start journalnode'
#ssh hadoop109 '/opt/module/HA/hadoop-2.7.2/sbin/hadoop-daemon.sh start journalnode'

#echo "==================关闭HDFS==========================="
ssh hadoop108 '/opt/module/HA/hadoop-2.7.2/sbin/stop-dfs.sh'

echo "================     正在启动Zookeeper               ==========="
    for i in hadoop108 hadoop109 hadoop110
        do
            ssh $i 'source /etc/profile ;cd /opt/module/zookeeper-3.4.10/bin/ ; ./zkServer.sh start'
            echo $i 执行完
    done
echo "================初始化HA在Zookeeper中状态=========================="
ssh hadoop108 '/opt/module/HA/hadoop-2.7.2/bin/hdfs zkfc -formatZK'

echo "================     正在启动HDFS                    ==========="
ssh hadoop108 '/opt/module/HA/hadoop-2.7.2/sbin/start-dfs.sh'

echo "==============启动DFSZK Failover Controller============="
ssh hadoop108 '/opt/module/HA/hadoop-2.7.2/sbin/hadoop-daemon.sh start zkfc'
ssh hadoop109 '/opt/module/HA/hadoop-2.7.2/sbin/hadoop-daemon.sh start zkfc'

echo "================     正在启动YARN                    ==========="
ssh hadoop108 '/opt/module/HA/hadoop-2.7.2/sbin/start-yarn.sh'
ssh hadoop109 '/opt/module/HA/hadoop-2.7.2/sbin/yarn-daemon.sh start resourcemanager'

echo "======================查看服务状态==================="

for i in hadoop108 hadoop109 hadoop110
do
    echo "===========        $i        ==========="
    ssh $i '/opt/module/jdk1.8.0_191/bin/jps'
done
echo "-------------------------架构-------------------------------- ------------"
echo "-----hadoop108------------hadoop109---------------hadoop110- -------------"
echo "-----NameNode-------------NameNode----------------------------------------"
echo "-----JournalNode----------JournalNode-------------JournalNode-------------"
echo "-----DataNode-------------DataNode----------------DataNode----------------"
echo "-----ZK-------------------ZK----------------------ZK----------------------"
echo "-----ResourceManager------ResourceManager---------------------------------"
echo "-----NodeManager----------NodeManager-------------NodeManager-------------"

==========================查看=================================

#!/bin/bash
for i in hadoop108 hadoop109 hadoop110
do
    echo "===========        $i        ==========="
    ssh $i '/opt/module/jdk1.8.0_191/bin/jps'
done

===============================================