基于HADOOP3.0+Centos7.0的yarn基础配置；
执行步骤：（1）配置集群yarn （2）启动、测试集群（3）在yarn上执行wordcount案例
一、配置yarn集群
1.配置yarn-env.sh
添加：export JAVA_HOME=/opt/module/jdk1.8.0_144
2.配置yarn-site.xml
<!--在etc/hadoop/yarn-site.xml文件中，修改检查虚拟内存的属性为false-->
<property>
<name>yarn.nodemanager.vmem-check-enabled</name>
<value>false</value>
</property>
<!--引入hadoop路径-->
<property>
<name>yarn.application.classpath</name>
<value>
/opt/hadoop-2.6.0/etc/hadoop,
/opt/hadoop-2.6.0/share/hadoop/common/*,
/opt/hadoop-2.6.0/share/hadoop/common/lib/*,
/opt/hadoop-2.6.0/share/hadoop/hdfs/*,
/opt/hadoop-2.6.0/share/hadoop/hdfs/lib/*,
/opt/hadoop-2.6.0/share/hadoop/mapreduce/*,
/opt/hadoop-2.6.0/share/hadoop/mapreduce/lib/*,
/opt/hadoop-2.6.0/share/hadoop/yarn/*,
/opt/hadoop-2.6.0/share/hadoop/yarn/lib/*
</value>
</property>
3.配置：mapred-env.sh
export JAVA_HOME=/opt/module/jdk1.8.0_144
4.配置mapred-site.xml
<!-- 指定mr运行在yarn上-->
<property>
<name>mapreduce.framework.name</name>
<value>yarn</value>
</property>
<property>
<name>yarn.application.classpath</name>
<value>
/opt/hadoop-2.6.0/etc/hadoop,
/opt/hadoop-2.6.0/share/hadoop/common/*,
/opt/hadoop-2.6.0/share/hadoop/common/lib/*,
/opt/hadoop-2.6.0/share/hadoop/hdfs/*,
/opt/hadoop-2.6.0/share/hadoop/hdfs/lib/*,
/opt/hadoop-2.6.0/share/hadoop/mapreduce/*,
/opt/hadoop-2.6.0/share/hadoop/mapreduce/lib/*,
/opt/hadoop-2.6.0/share/hadoop/yarn/*,
/opt/hadoop-2.6.0/share/hadoop/yarn/lib/*
</value>
</property>
二、启动集群
1.启动hdfs集群（namenode+datanode）
2.启动yarn集群（resourceManger+nodeManager）
sbin/start-yarn.sh or pasting
访问 http://192.168.1.101:8088/cluster
三、运行案例
1.删除文件系统上原来的output文件
hdfs dfs -rm -R /user/atguigu/output
2.执行mapreduce程序
hadoop jar share/hadoop/mapreduce/hadoop-mapreduce-examples-2.7.2.jar wordcount /user/atguigu/input /user/atguigu/output
3.查看运行结果
hdfs dfs -cat /user/atguigu/output/*