Hadoop HA 双namenode搭建

机器分布

hadoop1 192.168.56121

hadoop2 192.168.56122

hadoop3 192.168.56123

准备安装包

jdk-7u71-linux-x64.tar.gz

zookeeper-3.4.9.tar.gz

hadoop-2.9.2.tar.gz

把安装包上传到三台机器的/usr/local目录下并解压

配置hosts

echo "192.168.56.121 hadoop1" >> /etc/hosts
echo "192.168.56.122 hadoop2" >> /etc/hosts
echo "192.168.56.123 hadoop3" >> /etc/hosts

配置环境变量

/etc/profile

export HADOOP_PREFIX=/usr/local/hadoop-2.9.2
export JAVA_HOME=/usr/local/jdk1.7.0_71

部署zookeeper

创建zoo用户

useradd zoo
passwd zoo

修改zookeeper目录的属主为zoo

chown zoo:zoo -R /usr/local/zookeeper-3.4.9

修改zookeeper配置文件

到/usr/local/zookeeper-3.4.9/conf目录

cp zoo_sample.cfg zoo.cfg
vi zoo.cfg
tickTime=2000
initLimit=10
syncLimit=5
dataDir=/usr/local/zookeeper-3.4.9
clientPort=2181
server.1=hadoop1:2888:3888
server.2=hadoop2:2888:3888
server.3=hadoop3:2888:3888

创建myid文件放在/usr/local/zookeeper-3.4.9目录下，myid文件中只保存1-255的数字，与zoo.cfg中server.id行中的id相同。

hadoop1中myid为1

hadoop2中myid为2

hadoop3中myid为3

在三台机器启动zookeeper服务

[zoo@hadoop1 zookeeper-3.4.9]$ bin/zkServer.sh start

验证zookeeper

[zoo@hadoop1 zookeeper-3.4.9]$ bin/zkServer.sh status
ZooKeeper JMX enabled by default
Using config: /usr/local/zookeeper-3.4.9/bin/../conf/zoo.cfg
Mode: follower

配置Hadoop

创建用户

useradd hadoop
passwd hadoop

修改hadoop目录属主为hadoop

chmod hadoop:hadoop -R /usr/local/hadoop-2.9.2

创建目录

mkdir /hadoop1 /hadoop2 /hadoop3
chown hadoop:hadoop /hadoop1
chown hadoop:hadoop /hadoop2
chown hadoop:hadoop /hadoop3

配置互信

ssh-keygen
ssh-copy-id -i ~/.ssh/id_rsa.pub hadoop@hadoop1
ssh-copy-id -i ~/.ssh/id_rsa.pub hadoop@hadoop2
ssh-copy-id -i ~/.ssh/id_rsa.pub hadoop@hadoop3
#使用如下命令测试互信
ssh hadoop1 date
ssh hadoop2 date
ssh hadoop3 date

配置环境变量

/home/hadoop/.bash_profile

export PATH=$JAVA_HOME/bin:$HADOOP_PREFIX/bin:$HADOOP_PREFIX/sbin:$PATH

配置参数

etc/hadoop/hadoop-env.sh 

export JAVA_HOME=/usr/local/jdk1.7.0_71

etc/hadoop/core-site.xml

<!-- 指定hdfs的nameservice为ns -->
 <property>
      <name>fs.defaultFS</name>
      <value>hdfs://ns</value>
 </property>
 <!--指定hadoop数据临时存放目录-->
 <property>
      <name>hadoop.tmp.dir</name>
      <value>/usr/loca/hadoop-2.9.2/temp</value>
 </property>
 <property>
      <name>io.file.buffer.size</name>
      <value>4096</value>
 </property>
 <!--指定zookeeper地址-->
 <property>
      <name>ha.zookeeper.quorum</name>
      <value>hadoop1:2181,hadoop2:2181,hadoop3:2181</value>
 </property>

etc/hadoop/hdfs-site.xml

<!--指定hdfs的nameservice为ns，须要和core-site.xml中的保持一致 -->
  <property>
      <name>dfs.nameservices</name>
      <value>ns</value>
  </property>
  <!-- ns下面有两个NameNode，分别是nn1，nn2 -->
  <property>
     <name>dfs.ha.namenodes.ns</name>
     <value>nn1,nn2</value>
  </property>
  <!-- nn1的RPC通讯地址 -->
  <property>
     <name>dfs.namenode.rpc-address.ns.nn1</name>
     <value>hadoop1:9000</value>
  </property>
  <!-- nn1的http通讯地址 -->
  <property>
      <name>dfs.namenode.http-address.ns.nn1</name>
      <value>hadoop1:50070</value>
  </property>
  <!-- nn2的RPC通讯地址 -->
  <property>
      <name>dfs.namenode.rpc-address.ns.nn2</name>
      <value>hadoop2:9000</value>
  </property>
  <!-- nn2的http通讯地址 -->
  <property>
      <name>dfs.namenode.http-address.ns.nn2</name>
      <value>hadoop2:50070</value>
  </property>
  <!-- 指定NameNode的元数据在JournalNode上的存放位置 -->
  <property>
       <name>dfs.namenode.shared.edits.dir</name>
       <value>qjournal://hadoop1:8485;hadoop2:8485;hadoop3:8485/ns</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>/hadoop1/hdfs/journal</value>
  </property>
  <!-- 开启NameNode故障时自动切换 -->
  <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.ns</name>
          <value>org.apache.hadoop.hdfs.server.namenode.ha.ConfiguredFailoverProxyProvider</value>
  </property>
  <!-- 配置隔离机制，若是ssh是默认22端口，value直接写sshfence便可 -->
  <property>
           <name>dfs.ha.fencing.methods</name>
           <value>sshfence</value>
  </property>
  <!-- 使用隔离机制时须要ssh免登录 -->
  <property>
          <name>dfs.ha.fencing.ssh.private-key-files</name>
          <value>/home/hadoop/.ssh/id_rsa</value>
  </property>
  <property>
      <name>dfs.namenode.name.dir</name>
      <value>file:/hadoop1/hdfs/name,file:/hadoop2/hdfs/name</value>
  </property>
  <property>
      <name>dfs.datanode.data.dir</name>
      <value>file:/hadoop1/hdfs/data,file:/hadoop2/hdfs/data,file:/hadoop3/hdfs/data</value>
  </property>
  <property>
     <name>dfs.replication</name>
     <value>2</value>
  </property>
  <!-- 在NN和DN上开启WebHDFS (REST API)功能,不是必须 -->
  <property>
     <name>dfs.webhdfs.enabled</name>
     <value>true</value>
  </property>
  <property>
  <!-- List of permitted/excluded DataNodes.  -->
<name>dfs.hosts.exclude</name>
<value>/usr/local/hadoop-2.9.2/etc/hadoop/excludes</value>
</property>

​

etc/hadoop/mapred-site.xml

<property>
      <name>mapreduce.framework.name</name>
      <value>yarn</value>
  </property>
yarn-site.xml
  <!-- 指定nodemanager启动时加载server的方式为shuffle server -->
  <property>
          <name>yarn.nodemanager.aux-services</name>
          <value>mapreduce_shuffle</value>
   </property>
   <property>
          <name>yarn.nodemanager.aux-services.mapreduce_shuffle.class</name>
          <value>org.apache.hadoop.mapred.ShuffleHandler</value>
   </property>
   <!-- 指定resourcemanager地址 -->
   <property>
          <name>yarn.resourcemanager.hostname</name>
          <value>hadoop1</value>
    </property>

etc/hadoop/slaves

hadoop1
hadoop2
hadoop3

首次启动命令

1、首先启动各个节点的Zookeeper，在各个节点上执行以下命令：
bin/zkServer.sh start
2、在某一个namenode节点执行如下命令，创建命名空间
hdfs zkfc -formatZK
3、在每个journalnode节点用如下命令启动journalnode
sbin/hadoop-daemon.sh start journalnode
4、在主namenode节点格式化namenode和journalnode目录
hdfs namenode -format ns
5、在主namenode节点启动namenode进程
sbin/hadoop-daemon.sh start namenode
6、在备namenode节点执行第一行命令，这个是把备namenode节点的目录格式化并把元数据从主namenode节点copy过来，并且这个命令不会把journalnode目录再格式化了！然后用第二个命令启动备namenode进程！
hdfs namenode -bootstrapStandby
sbin/hadoop-daemon.sh start namenode
7、在两个namenode节点都执行以下命令
sbin/hadoop-daemon.sh start zkfc
8、在所有datanode节点都执行以下命令启动datanode
sbin/hadoop-daemon.sh start datanode

日常启停命令

#启动脚本，启动所有节点服务
sbin/start-dfs.sh
#停止脚本，停止所有节点服务
sbin/stop-dfs.sh验证

jps检查进程

​​http://192.168.56.122:50070​​

​​http://192.168.56.121:50070​​

测试文件上传下载

#创建目录
[hadoop@hadoop1 ~]$ hadoop fs -mkdir /test
#验证
[hadoop@hadoop1 ~]$ hadoop fs -ls /
Found 1 items
drwxr-xr-x   - hadoop supergroup          0 2019-04-12 12:16 /test    
#上传文件
[hadoop@hadoop1 ~]$ hadoop fs -put /usr/local/hadoop-2.9.2/LICENSE.txt /test
#验证
[hadoop@hadoop1 ~]$ hadoop fs -ls /test                                     
Found 1 items
-rw-r--r--   2 hadoop supergroup     106210 2019-04-12 12:17 /test/LICENSE.txt
#下载文件到/tmp
[hadoop@hadoop1 ~]$ hadoop fs -get /test/LICENSE.txt /tmp
#验证
[hadoop@hadoop1 ~]$ ls -l /tmp/LICENSE.txt 
-rw-r--r--. 1 hadoop hadoop 106210 Apr 12 12:19 /tmp/LICENSE.txt

参考：​​https://blog.csdn.net/Trigl/article/details/55101826​​