hadoop高可用分布式集群安装HDFS +yarm

一.  环境

准备4台linux虚拟机搭服务,1台pc机

三.  高可用分布式集群安装HDFS  50070

3.0部署图

 

 

 

 

 

 

NN-1

NN-2

DN

ZK

ZKFC

JNN

IP地址

Node01

*

 

 

 

*

*

192.168.21.148

Node02

 

*

*

*

*

*

192.168.21.147

Node03

 

 

*

*

 

*

192.168.21.143

Node04

 

 

*

*

 

 

192.168.21.146

 

 

 

 

 

 

 

 

3.1 hosts与profile配置

配置node01

配置node02

配置node03

配置node04

vim /etc/hosts
192.168.21.148 node01  hado
192.168.21.147 node02  hado
192.168.21.143 node03  hado
192.168.21.146 node04  hado
vim /etc/profile
export JAVA_HOME=/usr/local/jdk
export ZK_HOME=/usr/local/zookeeper
export HADOOP_HOME=/usr/local/hadoop-3.2.0-ha
PATH=$PATH::$JAVA_HOME/bin:$HADOOP_HOME/bin:$HADOOP_HOME/sbin:$HADOOP_HOME/lib:$ZK_HOME/bin

3.2 四台机器免登陆设置

配置node01

配置node02

配置node03

配置node04

ssh-keygen -t rsa

cat ~/.ssh/id_rsa.pub >> ~/.ssh/authorized_keys

chmod 700 ~/.ssh

chmod 600 ~/.ssh/authorized_keys

ssh-copy-id root@node02 

ssh-copy-id root@node03

ssh-copy-id root@node04

ssh-copy-id root@node01

ssh-copy-id root@node03

ssh-copy-id root@node04

ssh-copy-id root@node01

ssh-copy-id root@node02 

ssh-copy-id root@node04

ssh-copy-id root@node01

ssh-copy-id root@node02 

ssh-copy-id root@node03
           

报错

 

 

 

解决：

 

 

 

 

3.3解压

node01:

cd /usr/local/software
tar zxvf hadoop-3.2.0.tar.gz
mv hadoop-3.2.0  /usr/local/hadoop-3.2.0-ha

3.4配置修改

node01:

hadoop-env.sh

cd /usr/local/hadoop-3.2.0-ha/etc/hadoop

vim hadoop-env.sh

export JAVA_HOME=/usr/local/jdk

export HDFS_NAMENODE_USER=root

export HDFS_DATANODE_USER=root

export HDFS_SECONDARYNAMENODE_USER=root

export HDFS_ZKFC_USER=root

export HDFS_JOURNALNODE_USER=root

core-site.xml

cd /usr/local/hadoop-3.2.0-ha/etc/hadoop

vim core-site.xml

configuration中录入

  <property>

        <name>fs.defaultFS</name>

        <value>hdfs://mycluster </value>

    </property>

    <property>

        <name>hadoop.tmp.dir</name>

        <value>/var/sxt/hadoop/ha</value>

    </property>

   <property>

        <name>ha.zookeeper.quorum</name>

        <value>node02:2181,node03:2181,node04:2181</value>

    </property>

 

 

hdfs-site.xml

cd /usr/local/hadoop-3.2.0-ha/etc/hadoop
vim hdfs-site.xml
在configuration中录入
     <property>

        <name>dfs.replication</name>

        <value>1</value>

    </property>

    <property>

        <name>dfs.nameservices</name>

        <value>mycluster</value>

   </property>

   <property>

        <name>dfs.ha.namenodes.mycluster</name>

        <value>nn1,nn2</value>

   </property>

   <property>

        <name>dfs.namenode.rpc-address.mycluster.nn1</name>

        <value>node01:8020</value>

   </property>

   <property>

        <name>dfs.namenode.rpc-address.mycluster.nn2</name>

        <value>node02:8020</value>

   </property>

   <property>

         <name>dfs.namenode.http-address.mycluster.nn1</name>

   <value>node01:50070</value>

   </property>

   <property>

        <name>dfs.namenode.http-address.mycluster.nn2</name>

         <value>node02:50070</value>

   </property> <property>

        <name>dfs.namenode.shared.edits.dir</name>

        <value>qjournal://node01:8485;node02:8485;node03:8485/mycluster</value>

  </property>

  <property>

       <name>dfs.journalnode.edits.dir</name>

       <value>/var/sxt/hadoop/ha/jn</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</value>

  </property>

  <property>

       <name>dfs.ha.fencing.ssh.private-key-files</name>

       <value>/root/.ssh/id_dsa</value>

  </property>

 <property>

      <name>dfs.ha.automatic-failover.enabled</name>

      <value>true</value>

 </property>

workers

cd /usr/local/hadoop-3.2.0-ha/etc/hadoop

vim workers

node02

node03

node04

分发

cd /usr/local

scp -r hadoop-3.2.0-ha jdk node02:`pwd`

scp -r hadoop-3.2.0-ha jdk node03:`pwd`

scp -r hadoop-3.2.0-ha jdk node04:`pwd`

cd /etc

scp  profile node02:`pwd`

scp  profile node03:`pwd`

scp  profile node04:`pwd`

source /etc/profile 所有机器均执行

3.5 NameNode同步穿透

1>Node01格式化

Node01

Node02

Node03

cd /usr/local/hadoop-3.2.0-ha/bin

./hdfs --daemon start journalnode

cd /usr/local/hadoop-3.2.0-ha/bin

./hdfs namenode -format

 

 

2>查看Nod01版本

cd /var/sxt/hadoop/ha/dfs/name/current
cat V*

 

 

 

3>同步

启动node01的NameNode

Node01:
cd /usr/local/hadoop-3.2.0-ha/sbin

./hadoop-daemon.sh start namenode

Nod02同步

Node02:
cd /usr/local/hadoop-3.2.0-ha/bin

./hdfs namenode -bootstrapStandby

2>查看Nod02版本

cd /var/sxt/hadoop/ha/dfs/name/current

cat V*

 

 

 

 

3.6 zk格式化

node02,node03,node04机器上安装zookeeper集群，并启动。略

cd /usr/local/zookeeper/bin

./zkServer.sh start

Node01机器zk格式化

Node01:

cd /usr/local/hadoop-3.2.0-ha/bin

./hdfs zkfc -formatZK

 

3.7启动

Node01:

cd /usr/local/hadoop-3.2.0-ha/sbin

./start-dfs.sh

3.8登录验证

一台standby 一台active说明开启成功

 

 

 

进入zk客户端，注册节点位置如下图

 

 

 

3.9开机自启

Node01

Node02

Node03

Node04

vim /etc/rc.local    

cd /usr/local/hadoop-3.2.0-ha/sbin

./start-dfs.sh 

:wq

source  /etc/rc.local

chmod +x /etc/rc.local

vim /etc/rc.local    

cd /usr/local/zookeeper/bin

./zkServer.sh start

:wq

source  /etc/rc.local

chmod +x /etc/rc.local

四.  高可用分布式安装Yarn  8088

4.0部署图

 

 

 

NN-1

NN-2

DN

ZK

ZKFC

JNN

RS

NM

IP地址

Node01

*

 

 

 

*

*

 

 

192.168.21.148

Node02

 

*

*

*

*

*

 

*

192.168.21.147

Node03

 

 

*

*

 

*

*

*

192.168.21.143

Node04

 

 

*

*

 

 

*

*

192.168.21.146

4.1配置修改

hadoop-env.sh

cd /usr/local/hadoop-3.2.0-ha/etc/hadoop

vim hadoop-env.sh

export JAVA_HOME=/usr/local/jdk

export HDFS_NAMENODE_USER=root

export HDFS_DATANODE_USER=root

export HDFS_SECONDARYNAMENODE_USER=root

export HDFS_ZKFC_USER=root

export HDFS_JOURNALNODE_USER=root

 

export YARN_RESOURCEMANAGER_USER=root

export YARN_NODEMANAGER_USER=root

mapred-site.xml

cd /usr/local/hadoop-3.2.0-ha/etc/hadoop/mapred-site.xml

vim mapred-site.xml

configuration中录入

  <property>

    <name>mapreduce.framework.name</name>

    <value>yarn</value>

  </property>

 <property>

         <name>mapreduce.application.classpath</name>

         <value>

            /usr/local/hadoop-3.2.0-ha/etc/hadoop,

            /usr/local/hadoop-3.2.0-ha/share/hadoop/common/*,

            /usr/local/hadoop-3.2.0-ha/share/hadoop/common/lib/*,

            /usr/local/hadoop-3.2.0-ha/share/hadoop/hdfs/*,

            /usr/local/hadoop-3.2.0-ha/share/hadoop/hdfs/lib/*,

            /usr/local/hadoop-3.2.0-ha/share/hadoop/mapreduce/*,

            /usr/local/hadoop-3.2.0-ha/share/hadoop/mapreduce/lib/*,

            /usr/local/hadoop-3.2.0-ha/share/hadoop/yarn/*,

            /usr/local/hadoop-3.2.0-ha/share/hadoop/yarn/lib/*

         </value>

</property>

yarn-site.xml

cd /usr/local/hadoop-3.2.0-ha/etc/hadoop/

vim yarn-site.xml

configuration中录入

<!-- Site specific YARN configuration properties -->

<property>

    <name>yarn.nodemanager.aux-services</name>

    <value>mapreduce_shuffle</value>

</property>

<property>

    <name>yarn.nodemanager.env-whitelist</name>

    <value>JAVA_HOME,HADOOP_HOME</value>

</property>

<!--启用ResourceManager的高可用-->

<property>

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

  <value>true</value>

</property>

<!--指代ResourceManager HA的两台RM的逻辑名称 -->

<property>

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

  <value>rmhacluster1</value>

</property>

<!--指定该高可用ResourceManager下的两台ResourceManager的逻辑名称-->

<property>

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

  <value>rm1,rm2</value>

</property>

<!--指定第一台ResourceManager服务器所在的主机名称 -->

<property>

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

  <value>node03</value>

</property>

<property>

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

  <value>node04</value>

</property>

<!--指定resourcemanager的web服务器的主机名和端口号-->

<property>

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

  <value>node03:8088</value>

</property>

<property>

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

  <value>node04:8088</value>

</property>

<!--做ResourceManager HA故障切换用到的zookeeper集群地址 -->

<property>

  <name>yarn.resourcemanager.zk-address</name>

  <value>node02:2181,node03:2181,node04:2181</value>

</property>

分发

cd /usr/local/hadoop-3.2.0-ha/etc/hadoop/

scp -r yarn-site.xml  mapred-site.xml  hadoop-env.sh  node02:`pwd`

scp -r yarn-site.xml  mapred-site.xml  hadoop-env.sh  node03:`pwd`

scp -r yarn-site.xml  mapred-site.xml  hadoop-env.sh  node04:`pwd`

 

4.2启动

Node01:

cd /usr/local/hadoop-3.2.0-ha/sbin

./start-yarn.sh

4.3登录验证

http:192.168.143.8088

 工具下载

点击此处

posted @ 2020-09-14 21:20  咔咔kk  阅读(252)  评论(0)  编辑  收藏  举报