实现hadoop的HA配置

一、集群的规划

Zookeeper集群：

192.168.100.100 （master）
192.168.100.101 （hadoop01）
192.168.100.102 （hadoop02）

Hadoop集群：

192.168.100.100 （master）     NameNode1      ResourceManager1     Journalnode
192.168.100.101 （hadoop01）   NameNode2      ResourceManager2     Journalnode
192.168.100.102 （hadoop02）   DataNode1      NodeManager1
192.168.100.103 （hadoop03）   DataNode2      NodeManager2

二、准备工作

1、安装JDK
2、配置环境变量
3、配置免密码登录
4、配置主机名

三、配置Zookeeper（在master安装）

（1）tar -zxvf zookeeper-3.4.10.tar.gz -C /root/zookeeper-3.4.10
（2）/root/zookeeper-3.4.10/conf
（3）cp zoo_sample.cfg zoo.cfg
（4）vim zoo.cfg
（5）修改
dataDir=/root/zookeeper-3.4.10/zkData
###############cluster###############
server.0=master:2888:3888
server.1=hadoop01:2888:3888
server.2=hadoop02:2888:3888

（6）cd /root/zookeeper-3.4.10
（7）mkdir zkData
cd zkData/
touch myid
vim myid
0
（注意修改其他两台机器的id为1 2）
（8）发送zookeeper文件到其它机器

scp -r zookeeper-3.4.10/ hadoop01:/root/
scp -r zookeeper-3.4.10/ hadoop02:/root/

四、安装Hadoop集群（在master上安装）

1、修改hadoo-env.sh
export JAVA_HOME=/usr/local/lib/jdk1.8.0_181

2、修改core-site.xml

<configuration>
    <!-- 指定hdfs的nameservice为ns1 -->
    <property>
            <name>fs.defaultFS</name>
            <value>hdfs://ns1</value>
    </property>

    <!-- 指定hadoop临时目录 -->
    <property>
            <name>hadoop.tmp.dir</name>
            <value>/root/hadoop-2.7.3/dfs/data/tmp</value>
    </property>

    <!-- 指定zookeeper地址 -->
    <property>
            <name>ha.zookeeper.quorum</name>
            <value>master:2181,hadoop01:2181,hadoop02:2181</value>
    </property>
</configuration>

3、修改hdfs-site.xml（配置这个nameservice中有几个namenode）

<configuration> 
    <!--指定hdfs的nameservice为ns1，需要和core-site.xml中的保持一致 -->
    <property>
        <name>dfs.nameservices</name>
        <value>ns1</value>
    </property>

    <!-- ns1下面有两个NameNode，分别是nn1，nn2 -->
    <property>
        <name>dfs.ha.namenodes.ns1</name>
        <value>nn1,nn2</value>
    </property>

    <!-- nn1的RPC通信地址 -->
    <property>
        <name>dfs.namenode.rpc-address.ns1.nn1</name>
        <value>master:9000</value>
    </property>
    
    <!-- nn1的http通信地址 -->
    <property>
        <name>dfs.namenode.http-address.ns1.nn1</name>
        <value>master:50070</value>
    </property>

    <!-- nn2的RPC通信地址 -->
    <property>
        <name>dfs.namenode.rpc-address.ns1.nn2</name>
        <value>hadoop01:9000</value>
    </property>
    
    <!-- nn2的http通信地址 -->
    <property>
        <name>dfs.namenode.http-address.ns1.nn2</name>
        <value>hadoop01:50070</value>
    </property>

    <!-- 指定NameNode的日志在JournalNode上的存放位置 -->
    <property>
        <name>dfs.namenode.shared.edits.dir</name>
        <value>qjournal://master:8485;hadoop01:8485;/ns1</value>
    </property>
    
    <!-- 指定JournalNode在本地磁盘存放数据的位置 -->
    <property>
        <name>dfs.journalnode.edits.dir</name>
        <value>/root/hadoop-2.7.3/dfs/data</value>
    </property>

    <!-- 开启NameNode失败自动切换 -->
    <property>
        <name>dfs.ha.automatic-failover.enabled</name>
        <value>true</value>
    </property>

    <!-- 配置失败自动切换实现方式 -->
    <property>
        <name>dfs.client.failover.proxy.provider.ns1</name>
        <value>org.apache.hadoop.hdfs.server.namenode.ha.ConfiguredFailoverProxyProvider</value>
    </property>

    <!-- 配置隔离机制方法，多个机制用换行分割，即每个机制暂用一行-->
    <property>
        <name>dfs.ha.fencing.methods</name>
        <value>
            sshfence
            shell(/bin/true)
        </value>
    </property>

    <!-- 使用sshfence隔离机制时需要ssh免登陆 -->
    <property>
        <name>dfs.ha.fencing.ssh.private-key-files</name>
        <value>/root/.ssh/id_rsa</value>
    </property>

    <!-- 配置sshfence隔离机制超时时间 -->
    <property>
        <name>dfs.ha.fencing.ssh.connect-timeout</name>
        <value>30000</value>
    </property>
</configuration>

4、修改mapred-site.xml

	<configuration>
		<property>
				<name>mapreduce.framework.name</name>
				<value>yarn</value>
		</property>
	</configuration>

5、修改yarn-site.xml

<configuration>
    <!-- 开启RM高可靠 -->
    <property>
       <name>yarn.resourcemanager.ha.enabled</name>
       <value>true</value>
    </property>

    <!-- 指定RM的cluster id -->
    <property>
       <name>yarn.resourcemanager.cluster-id</name>
       <value>yrc</value>
    </property>

    <!-- 指定RM的名字 -->
    <property>
       <name>yarn.resourcemanager.ha.rm-ids</name>
       <value>rm1,rm2</value>
    </property>

    <!-- 分别指定RM的地址 -->
    <property>
       <name>yarn.resourcemanager.hostname.rm1</name>
       <value>master</value>
    </property>
    
    <property>
       <name>yarn.resourcemanager.hostname.rm2</name>
       <value>hadoop01</value>
    </property>

    <!-- 指定zk集群地址 -->
    <property>
       <name>yarn.resourcemanager.zk-address</name>
       <value>master:2181,hadoop01:2181,hadoop02:2181</value>
    </property>

    <property>
       <name>yarn.nodemanager.aux-services</name>
       <value>mapreduce_shuffle</value>
    </property>
</configuration>

6、修改slaves

hadoop02
hadoop03

7、将配置好的hadoop拷贝到其他节点

	scp -r /root/hadoop-2.7.3/ hadoop01:/root/
	scp -r /root/hadoop-2.7.3/ hadoop02:/root/
	scp -r /root/hadoop-2.7.3/ hadoop03:/root/

五、启动Zookeeper集群

start-zookeeper.sh

六、在master和hadoop01上启动journalnode

hadoop-daemon.sh start journalnode

七、格式化HDFS（在master上执行）

1.hdfs namenode -format
2. 将/root/hadoop-2.7.3/dfs/data/tmp拷贝到hadoop01的/root/hadoop-2.7.3/dfs/data/tmp下
hadoop01上执行删除这个目录下的数据：
rm -rf dfs/*
拷贝master的dfs目录下文件：
cd /root/hadoop-2.7.3
scp -r dfs/* hadoop01:/root/hadoop-2.7.3/dfs/

3. 格式化zookeeper
   hdfs zkfc -formatZK
   日志：19/02/27 19:14:05 INFO ha.ActiveStandbyElector: Successfully created /hadoop-ha/ns1 in ZK.

八、在master上启动Hadoop集群

start-all.sh（或者start-dfs.sh 和 start-yarn.sh 单独使用）

日志：
	Starting namenodes on [master hadoop01]
    master: starting namenode, logging to /root/hadoop-2.7.3/logs/hadoop-root-namenode-master.out
    hadoop01: starting namenode, logging to /root/hadoop-2.7.3/logs/hadoop-root-namenode-hadoop01.out
    hadoop03: starting datanode, logging to /root/hadoop-2.7.3/logs/hadoop-root-datanode-hadoop03.out
    hadoop02: starting datanode, logging to /root/hadoop-2.7.3/logs/hadoop-root-datanode-hadoop02.out
    Starting journal nodes [master hadoop01 ]
    master: journalnode running as process 5794. Stop it first.
    hadoop01: journalnode running as process 2620. Stop it first.
    Starting ZK Failover Controllers on NN hosts [master hadoop01]
    master: starting zkfc, logging to /root/hadoop-2.7.3/logs/hadoop-root-zkfc-master.out
    hadoop01: starting zkfc, logging to /root/hadoop-2.7.3/logs/hadoop-root-zkfc-hadoop01.out
    starting yarn daemons
    starting resourcemanager, logging to /root/hadoop-2.7.3/logs/yarn-root-resourcemanager-master.out
    hadoop03: starting nodemanager, logging to /root/hadoop-2.7.3/logs/yarn-root-nodemanager-hadoop03.out
    hadoop02: starting nodemanager, logging to /root/hadoop-2.7.3/logs/yarn-root-nodemanager-hadoop02.out

九、hadoop01上的ResourceManager需要单独启动

命令：yarn-daemon.sh start resourcemanager

十、HA实现成功的标志

在master的50070端口查看是active状态：

在hadoop01的50070端口查看是standby状态：