centos7中Yarn分布式安装

hadoop11上边启动namenode/resourceManager
启动HDFS集群  start-dfs.sh (在namenode进程所在主机运行该命令)
启动Yarn集群  start-yarn.sh (在resourceManager进程所在主机运行该命令)
如果在hadoop13上边启动yarn，注意配置hadoop13到集群节点的免密登录

hadoop11
        NameNode
        DataNode
        ResourceManager
		NodeManager
		JobHistoryServer 

hadoop12
        SecondaryNameNode 
        DataNode
        NodeManager
        
hadoop13
        DataNode
        NodeManager

# 保证HDFS分布式集群搭建环境确保正确。

1. hadoop-env.sh
export YARN_RESOURCEMANAGER_USER=root
export YARN_NODEMANAGER_USER=root

2. mapred-site.xml
	<property>
        <name>mapreduce.framework.name</name>
        <value>yarn</value>
    </property>
    
    <!-- 历史服务器端地址 -->
    <property>
      <name>mapreduce.jobhistory.address</name>
      <value>hadoop11:10020</value>
    </property>

    <!-- 历史服务器 web 端地址 -->
    <property>
      <name>mapreduce.jobhistory.webapp.address</name>
      <value>hadoop11:19888</value>
    </property>

3. yarn-site.xml
	<property>
        <name>yarn.nodemanager.aux-services</name>
        <value>mapreduce_shuffle</value>
    </property>
	<!-- 配置resourcemanager的主机ip -->
    <property>
        <name>yarn.resourcemanager.hostname</name>
        <value>hadoop11</value>
    </property>
    
    <property>
		  <name>yarn.application.classpath</name>
		  <value>
				/opt/installs/hadoop3.1.4/etc/hadoop,
				/opt/installs/hadoop3.1.4/share/hadoop/common/*,
				/opt/installs/hadoop3.1.4/share/hadoop/common/lib/*,
				/opt/installs/hadoop3.1.4/share/hadoop/hdfs/*,
				/opt/installs/hadoop3.1.4/share/hadoop/hdfs/lib/*,
				/opt/installs/hadoop3.1.4/share/hadoop/mapreduce/*,
				/opt/installs/hadoop3.1.4/share/hadoop/mapreduce/lib/*,
				/opt/installs/hadoop3.1.4/share/hadoop/yarn/*,
				/opt/installs/hadoop3.1.4/share/hadoop/yarn/lib/*
		  </value>
    </property>	
    
    <!-- 开启日志聚合 -->
    <property>
      <name>yarn.log-aggregation-enable</name>
      <value>true</value>
    </property>

    <!-- 设置日志保存时间 7 天 -->
    <property>
      <name>yarn.log-aggregation.retain-seconds</name>
      <value>604800</value>
    </property>

4. workers
	配置nodemanager(从机)所在的ip。
	配置datanode(从机)所在的ip
# 2：同步该配置到其他节点服务器上。
[root@hadoop11 etc]# scp -r hadoop/ root@hadoop12:/opt/installs/hadoop2.9.2/etc/
[root@hadoop11 etc]# scp -r hadoop/ root@hadoop13:/opt/installs/hadoop2.9.2/etc/

myscp分发脚本

# 3：启动yarn集群
# 在namenode所在主机上
1. 启动HDFS集群
	start-dfs.sh
# 在Resourcemanager所在主机上
2. 启动yarn集群
	start-yarn.sh
	mapred --daemon start historyserver

# 4：验证
1. jps

[root@hadoop11 ~]# jps
6160 DataNode
6513 ResourceManager
6614 NodeManager
6056 NameNode
6349 SecondaryNameNode
6831 Jps
25199 JobHistoryServer

2. 访问yarn的资源调度器web网页。
	http://主节点ResourceManager节点的ip:8088

# 关闭集群
1. 先关闭yarn和历史日志服务器    
	stop-yarn.sh
	mapred --daemon stop historyserver
2. 在关闭hdfs
	stop-hdfs.sh