三台linux，yum和网络都配置好
yum和网络配置点这里
jdk和hadoop的安装包上传上来
master下执行
hostnamectl set-hostname master bash
slave1下执行
hostnamectl set-hostname slave1 bash
slave2下执行
hostnamectl set-hostname slave2 bash2.主机映射
三台都要执行
vi /etc/hosts
末尾写入
192.168.26.148 master 192.168.26.149 slave1 192.168.26.150 slave23.ssh免密
仅master执行
master输入
ssh-keygen -t rsa
三次回车出现密文如下（大概这个样子，不完全一样）
+--[ RSA 2048]----+ | .o.+. .. | |. . o.. .E. | |.o.++ . * . | |..+o.. + = | | +. S+ | | .. . . | | . | | | | | +-----------------+
ssh-copy-id -i /root/.ssh/id_rsa.pub master ssh-copy-id -i /root/.ssh/id_rsa.pub slave1 ssh-copy-id -i /root/.ssh/id_rsa.pub slave2
依次输入 yes和root 用户的密码
依次验证
ssh master ssh slave1 ssh slave2
登录一次就及时退出一次（exit）
4.关闭防火墙三台都要执行
systemctl stop firewalld systemctl disable firewalld
查看防火墙状态
systemctl status firewalld5.安装jdk
仅master执行
解压缩
tar -zxvf jdk-8u152-linux-x64.tar.gz -C /usr/
改个好名字，直接改成jdk
mv /usr/jdk1.8.0_261/ /usr/jdk6.安装hadoop
仅master执行
解压缩
tar -zxvf hadoop-3.1.3.tar.gz -C /usr/
改个好名字，直接改成hadoop
mv /usr/hadoop-3.1.3/ /usr/hadoop7.hadoop配置
仅master执行
①进入目录
cd /usr/hadoop/etc/hadoop
② core-site.xml
vi core-site.xml
fs.defaultFS hdfs://master:9000 HDFS的URI hadoop.tmp.dir file:/usr/hadoop/tmp 节点上本地的hadoop临时文件夹
③ hadoop-env.sh
vi hadoop-env.sh
export JAVA_HOME=/usr/jdk
④ hdfs-site.xml
vi hdfs-site.xml
dfs.namenode.http-address master:9870 dfs.namenode.name.dir file:/usr/hadoop/hdfs/name namenode上存储hdfs名字空间元数据 dfs.datanode.data.dir file:/usr/hadoop/hdfs/data datanode上数据块的物理存储位置 dfs.replication 3 副本个数，默认是3,应小于datanode机器数量
⑤ mapred-site.xml
vi mapred-site.xml
mapreduce.framework.name yarn 指定mapreduce使用yarn框架 yarn.app.mapreduce.am.env HADOOP_MAPRED_HOME=${HADOOP_HOME} mapreduce.map.env HADOOP_MAPRED_HOME=${HADOOP_HOME} mapreduce.reduce.env HADOOP_MAPRED_HOME=${HADOOP_HOME}
⑥ yarn-site.xml
vi yarn-site.xml
yarn.resourcemanager.hostname master 指定resourcemanager所在的hostname yarn.nodemanager.aux-services mapreduce_shuffle NodeManager上运行的附属服务。需配置成mapreduce_shuffle，才可运行MapReduce程序
⑦ yarn-env.sh
vi yarn-env.sh
export JAVA_HOME=/usr/jdk
⑧ workers
删除原有内容
添加
vi workers
slave1 slave2
⑨配置环境变量
vi /etc/profile
写入
#jdk export JAVA_HOME=/usr/jdk export PATH=$PATH:$JAVA_HOME/bin #hadoop export HADOOP_HOME=/usr/hadoop export PATH=$HADOOP_HOME/bin:$PATH:$HADOOP_HOME/sbin export HDFS_NAMENODE_USER=root export HDFS_DATANODE_USER=root export HDFS_SECONDARYNAMENODE_USER=root export YARN_RESOURCEMANAGER_USER=root export YARN_NODEMANAGER_USER=root
刷新环境变量
source /etc/profile8.分发文件
jdk
scp -r /usr/jdk slave1:/usr/ scp -r /usr/jdk slave2:/usr/
hadoop
先创建文件夹
mkdir /usr/hadoop/tmp mkdir -p /usr/hadoop/hdfs/name mkdir /usr/hadoop/hdfs/data
scp -r /usr/hadoop slave1:/usr/ scp -r /usr/hadoop slave2:/usr/
环境变量
scp -r /etc/profile slave1:/etc/profile scp -r /etc/profile slave2:/etc/profile
刷新环境变量
source /etc/profile9.格式化
hdfs namenode -format10.启动集群
start-all.sh11.查看集群状态
①jps
master
NameNode SecondaryNameNode Jps ResourceManager
slave
Jps DataNode NodeManager
②web
http://192.168.26.148:9870 http://192.168.26.148:808812.执行MapReduce词频统计任务
本地创建文件
cd /root vi 1.txt
写入
Give me the strength lightly to bear my joys and sorrows.Give me the strength to make my love fruitful in service.Give me the strength never to disown the poor or bend my knees before insolent might.Give me the strength to raise my mind high above daily trifles.And give me the strength to surrender my strength to thy will with love.
创建文件夹并上传文件
hadoop fs -mkdir /input hadoop fs -put /root/1.txt /input
执行命令
cd /usr/hadoop/share/hadoop/mapreduce hadoop jar hadoop-mapreduce-examples-3.1.3.jar wordcount /input/1.txt /output
此处的hadoop-mapreduce-examples-3.1.3.jar 根据实际版本填写
查看结果
hadoop fs -ls /output hadoop fs -cat /output/part-r-00000
如下
Give 1 above 1 and 1 bear 1 before 1 bend 1 daily 1 disown 1 fruitful 1 give 1 high 1 in 1 insolent 1 joys 1 knees 1 lightly 1 love 1 love. 1 make 1 me 5 might.Give 1 mind 1 my 5 never 1 or 1 poor 1 raise 1 service.Give 1 sorrows.Give 1 strength 6 surrender 1 the 6 thy 1 to 6 trifles.And 1 will 1 with 1
