Hadoop HA一键启动，一键关闭脚本

本脚本实现一个简单的三台机器分布式集群环境下的Hadoop一键启动。

1：在目录 /usr/local/bin下 touch xhp.sh

2:   chmod 777 xhp.sh 

3:   编辑 xhp.sh

ZK:Zookeeper  ZKFC:Zookeeper FailoverController  JN:JournalNodes   RM:Yarn ResourceManager   DM:Yarn ApplicationManager(DataManager)

ZKFC 必须跟namenode在一起（hearbeat） DM 最好跟DataNode放一起,JN和ZK必须为奇数个(3,5,7)

机器编号	Namenode	DateNode	Zookeeper	ZKFC	JN	RM	DM
hadoop202	√	√	√	√	√		√
hadoop203	√	√	√	√	√	√	√
hadoop204		√	√		√	√	√

#判断用户是否传参
if [ $# -ne 1 ];then
    echo "无效参数，用法为: $0  {start|stop}"
    exit
fi
#获取用户输入的命令
cmd=$1
#定义函数功能
function hadoopManger(){
    case $cmd in
    start)
        echo "启动服务"        
        remoteExecutionstart
        ;;
    stop)
        echo "停止服务"
        remoteExecutionstop
        ;;
    *)
        echo "无效参数，用法为: $0  {start|stop}"
        ;;
    esac
}



#启动HADOOP
function remoteExecutionstart(){
    #zookeeper
    echo ==========   ZooKeeper   ================
    for (( i=202 ; i<=204 ; i++ )) ; 
    do
            tput setaf 2
            echo ========== 启动 hadoop${i} zookeeper  $1 ================
            tput setaf 9
            ssh hadoop${i}  "source /etc/profile ; /opt/module/zookeeper-3.4.10/bin/zkServer.sh start"
    done
    
    #journalnode 
    echo ==========   Journalnode   ================
    for (( i=202 ; i<=204 ; i++ )) ; 
    do
            tput setaf 2
            echo ==========启动  hadoop${i} journalnode  $1 ================
            tput setaf 9
            ssh hadoop${i}  "source /etc/profile ; /opt/module/hadoop-2.7.2/sbin/hadoop-daemon.sh start journalnode"
    done
    
    echo "启动HDFS"        
    ssh hadoop202  "source /etc/profile ; /opt/module/hadoop-2.7.2/sbin/start-dfs.sh"
    
    echo "启动YARM"        
    ssh hadoop202  "source /etc/profile ; /opt/module/hadoop-2.7.2/sbin/start-yarn.sh"
    
    echo "启动ResourceManager"        
    ssh hadoop203  "source /etc/profile ; /opt/module/hadoop-2.7.2/sbin/yarn-daemon.sh start resourcemanager"
    
}


#关闭HADOOP
function remoteExecutionstop(){
    
    echo "关闭ResourceManager"        
    ssh hadoop203  "source /etc/profile ; /opt/module/hadoop-2.7.2/sbin/yarn-daemon.sh stop resourcemanager"
    
    echo "关闭YARM"        
    ssh hadoop202  "source /etc/profile ; /opt/module/hadoop-2.7.2/sbin/stop-yarn.sh"
    
    echo "关闭HDFS"        
    ssh hadoop202  "source /etc/profile ; /opt/module/hadoop-2.7.2/sbin/stop-dfs.sh"    
    
    #journalnode
    echo ==========   Journalnode   ================
    for (( i=202 ; i<=204 ; i++ )) ; 
    do
            tput setaf 2
            echo ==========关闭  hadoop${i} journalnode  $1 ================
            tput setaf 9
            ssh hadoop${i}  "source /etc/profile ; /opt/module/hadoop-2.7.2/sbin/hadoop-daemon.sh stop journalnode"
    done
    
    #zookeeper
    echo ==========   ZooKeeper   ================
    for (( i=202 ; i<=204 ; i++ )) ; 
    do
            tput setaf 2
            echo ========== 关闭 hadoop${i} zookeeper  $1 ================
            tput setaf 9
            ssh hadoop${i}  "source /etc/profile ; /opt/module/zookeeper-3.4.10/bin/zkServer.sh stop"
    done
    
}
#调用函数
hadoopManger

 注意:Journalnode 的启动关闭脚本可以不需要.ZKFC 的启动关闭脚本也不需要。因为

the start-dfs.sh script will now automatically start a ZKFC daemon on any machine that runs a NameNode. When the ZKFCs start, they will automatically select one of the NameNodes to become active.

Journalnode 也会被自动启动.

运行脚本:查询启动结果

 web 端查看

http://hadoop202:50070/

 http://hadoop203:8088/