Hadoop的HDFS HA搭建

重新解压安装Hadoop-2.7.2到/opt/ha上

hadoop272etchadoop_5">进入/opt/ha/hadoop-2.7.2/etc/hadoop配置

hadoopenvsh_7">hadoop-env.sh

export JAVA_HOME=/opt/module/jdk1.8.0_144

core-site.xml

<configuration>
<!-- 指定HDFS中NameNode的地址 -->
<property>
	<name>fs.defaultFS</name>
	<value>hdfs://mycluster</value>
</property>
<!-- 指定Hadoop运行时产生文件的存储目录 -->
<property>
	<name>hadoop.tmp.dir</name>
	<value>/opt/ha/hadoop-2.7.2/data/tmp</value>
</property>
<property>
	<name>ha.zookeeper.quorum</name>
	<value>server1:2181,server2:2181,server3:2181</value>
</property>
</configuration>

hdfssitexml_34">hdfs-site.xml

<configuration>
<!-- 关闭权限检查-->
<property>
  <name>dfs.permissions.enable</name>
  <value>false</value>
</property>
<!-- 副本数 -->
<property>
  <name>dfs.replication</name>
  <value>3</value>
</property>
<!-- 指定Hadoop辅助名称节点主机配置 -->
<property>
  <name>dfs.namenode.secondary.http-address</name>
  <value>server3:50090</value>
</property>
<!-- 完全分布式集群名称-->
<property>
  <name>dfs.nameservices</name>
  <value>mycluster</value>
</property>
<!-- nameservice ID-->
<property>
  <name>dfs.ha.namenodes.mycluster</name>
  <value>nn1,nn2</value>
</property>
<!-- namenode address in fs.defaultFS of core-site-->
<property>
  <name>dfs.namenode.rpc-address.mycluster.nn1</name>
  <value>server1:9000</value>
</property>
<property>
  <name>dfs.namenode.rpc-address.mycluster.nn2</name>
  <value>server2:9000</value>
</property>
<!--namenode HTTP address-->
<property>
  <name>dfs.namenode.http-address.mycluster.nn1</name>
  <value>server1:50070</value>
</property>
<property>
  <name>dfs.namenode.http-address.mycluster.nn2</name>
  <value>server2:50070</value>
</property>
<!--the URI which identifies the group of JNs where the NameNodes will write/read edits-->
<property>
  <name>dfs.namenode.shared.edits.dir</name>
  <value>qjournal://server1:8485;server2:8485;server3:8485/mycluster</value>
</property>
<!--the Java class that HDFS clients use to contact the Active NameNode-->
<property>
  <name>dfs.client.failover.proxy.provider.mycluster</name>
  <value>org.apache.hadoop.hdfs.server.namenode.ha.ConfiguredFailoverProxyProvider</value>
</property>
<!--a list of scripts or Java classes which will be used to fence the Active NameNode during a failover-->
<property>
  <name>dfs.ha.fencing.methods</name>
  <value>sshfence</value>
</property>
<!--SSH to the Active NameNode and kill the process-->
<property>
  <name>dfs.ha.fencing.ssh.private-key-files</name>
  <value>/home/server/.ssh/id_rsa</value>
</property>
<property>
      <name>dfs.ha.fencing.methods</name>
      <value>sshfence</value>
</property>
<property>
  <name>dfs.ha.fencing.ssh.connect-timeout</name>
  <value>30000</value>
</property>
<property>
  <name>dfs.journalnode.edits.dir</name>
  <value>/opt/ha/hadoop-2.7.2/data/jn</value>
</property>
<property>
  <name>dfs.ha.automatic-failover.enabled</name>
  <value>true</value>
</property>
</configuration>

重新配置/etc/profile的Hadoop位置，并重启

这里需要注意，不重启仅仅只是source /etc/profile是不会生效的，你会发现NameNode在格式化的时候，还是格式化你原来的Hadoop

测试

1. 三台虚拟机都输入以下指令

sbin/hadoop-daemon.sh start journalnode

2. 第一个namenode的节点输入

bin/hdfs namenode -format
sbin/hadoop-daemon.sh start namenode

3. 第二个namenode的节点输入

bin/hdfs namenode -bootstrapStandby
sbin/hadoop-daemon.sh start namenode

4. 浏览器输入以下url，查看50070端口

   name1的ip:50070

   name2的ip:50070

5. 在第一个namenode节点上启动datanode

sbin/hadoop-daemons.sh start datanode

6. 将namenode转换为active

bin/hdfs haadmin -transitionToActive nn1

7. 查看状态

bin/hdfs haadmin -getServiceState nn1

测试没有问题，就开始启动

1. 关闭所有HDFS服务，namenode关不掉就kill -9 进程号杀死

sbin/stop-dfs.sh

2. 启动zookeeper

bin/zkServer.sh start

3. 初始化HA在zookeeper中的状态

bin/hdfs zkfc -formatZK

4. 启动HDFS服务

sbin/start-dfs.sh