Linux脚本，一键关闭启动hadoop中相关进程，启动MapReduce程序

废话

最近学习hadoop方面的知识，每次开启Linux时都要启动hadoop相应的四个进程

用命令行打着实有点麻烦，而且有时一个不小心打错字母，还得从头来一遍

于是想了下面这几个脚本文件，帮助启动关闭hadoop相关进程

建议新手还是要多去练一练

废话不多说，上代码

附：代码有点不规范，请见谅

start-hadoop.sh：启动相关hadoop相关进程

#!/bin/sh
#为防止出错，命令都写为了绝对路径
namenode=`/opt/modules/hadoop-2.7.3/sbin/hadoop-daemon.sh start namenode`
datanode=`/opt/modules/hadoop-2.7.3/sbin/hadoop-daemon.sh start datanode`
resourcemanager=`/opt/modules/hadoop-2.7.3/sbin/yarn-daemon.sh start resourcemanager`
nodemanager=`/opt/modules/hadoop-2.7.3/sbin/yarn-daemon.sh start nodemanager`
histoerserver=`/opt/modules/hadoop-2.7.3/sbin/mr-jobhistory-daemon.sh start historyserver`
echo $namenode
echo $datanode
echo $resourcemanager
echo $nodemanager
echo $histoerserver

stop-hadoop.sh：关闭相关hadoop进程

#!/bin/sh
#为防止出错，命令都写为了绝对路径
namenode=`/opt/modules/hadoop-2.7.3/sbin/hadoop-daemon.sh stop namenode`
datanode=`/opt/modules/hadoop-2.7.3/sbin/hadoop-daemon.sh stop datanode`
resourcemanager=`/opt/modules/hadoop-2.7.3/sbin/yarn-daemon.sh stop resourcemanager`
nodemanager=`/opt/modules/hadoop-2.7.3/sbin/yarn-daemon.sh stop nodemanager`
histoerserver=`/opt/modules/hadoop-2.7.3/sbin/mr-jobhistory-daemon.sh stop historyserver`
shutdowm=`sudo halt`
echo $namenode
echo $datanode
echo $resourcemanager
echo $nodemanager
echo $histoerserver
#关机命令，可以根据需要决定是否添加
#echo $shutdown

start-mapreduce-task.sh：启动MapReduce任务

#!/bin/sh
#输入运行jar包
echo "pleasd enter jar"
read jar
#输入运行主类
echo "please enter main class"
read mainclass
#输入数据源路径
echo "please enter source data"
read src
#输入结果存放位置
echo "please enter target file"
read target
#定义执行命令
order=`/opt/modules/hadoop-2.7.3/bin/yarn jar $jar $mainclass $src $target`
echo $order