HDFS的NameNode的NameNode的HA配置
一,zookeeper的安装(略)。
启动Zookeeper。
二,Hadoop配置文件修改。
修改core-site.xml
<configuration> <property> <name>fs.defaultFS</name> <value>hdfs://myCluster</value> <description>注意:myhadoop为集群的逻辑名,最好与hdfs-site.xml中的dfs.nameservices配置一致!否则在使用hdfs命令的时候,后面跟的路径就要指定全路径.如果不一致,那么执行hdfs dfs -ls /就会报错(java.net.UnknownHostException: ××××)。但可以通过这样的命令来执行:hdfs dfs -ls hdfs://myCluster </description> </property> <property> <name>ha.zookeeper.quorum</name> <value>trainingHadoop01:2181,trainingHadoop02:2181,trainingHadoop03:2181,trainingHadoop04:2181</value> <description>各个ZK节点的IP/host,及客户端连接ZK的端口,该端口需与zoo.cfg中的clientPort一致!</description> </property>
</configuration>
<configuration>
<property>
<name>fs.defaultFS</name>
<value>hdfs://myCluster</value>
<description>注意:myhadoop为集群的逻辑名,最好与hdfs-site.xml中的dfs.nameservices配置一致!否则在使用hdfs命令的时候,后面跟的路径就要指定全路径.如果不一致,那么执行hdfs
dfs -ls /就会报错(java.net.UnknownHostException: ××××)。但可以通过这样的命令来执行:hdfs
dfs -ls hdfs://myCluster
</description>
</property>
<property>
<name>ha.zookeeper.quorum</name>
<value>trainingHadoop01:2181,trainingHadoop02:2181,trainingHadoop03:2181,trainingHadoop04:2181</value>
<description>各个ZK节点的IP/host,及客户端连接ZK的端口,该端口需与zoo.cfg中的clientPort一致!</description>
</property>
</configuration>
修改hdfs-site.xml
<property>
<name>dfs.ha.automatic-failover.enabled</name>
<value>true</value>
<description>
Whether automatic failover is enabled. See the HDFS High
Availability documentation for details on automatic HA
configuration.
</description>
</property>
<property>
<name>dfs.nameservices</name>
<value>myCluster</value>
<description>
Comma-separated list of nameservices.
</description>
</property>
<property>
<name>dfs.ha.namenodes.myCluster</name>
<value>nn1,nn2</value>
<description>
The prefix for a given nameservice, contains a
comma-separated
list of namenodes for a given nameservice (eg
EXAMPLENAMESERVICE).
</description>
</property>
<property>
<name>dfs.namenode.rpc-address.myCluster.nn1</name>
<value>trainingHadoop01:8020</value>
</property>
<property>
<name>dfs.namenode.rpc-address.myCluster.nn2</name>
<value>trainingHadoop02:8020</value>
</property>
<property>
<name>dfs.namenode.http-address.myCluster.nn1</name>
<value>trainingHadoop01:50070</value>
</property>
<property>
<name>dfs.namenode.http-address.myCluster.nn2</name>
<value>trainingHadoop02:50070</value>
</property>
<property>
<name>dfs.namenode.servicerpc-address.myCluster.nn1</name>
<value>trainingHadoop01:53310</value>
</property>
<property>
<name>dfs.namenode.servicerpc-address.myCluster.nn2</name>
<value>trainingHadoop02:53310</value>
</property>
<property>
<name>dfs.namenode.shared.edits.dir</name>
<value>qjournal://trainingHadoop02:8485;trainingHadoop03:8485;trainingHadoop04:8485/myCluster
</value>
</property>
<property>
<name>dfs.ha.fencing.methods</name>
<value>shell(/bin/true)</value>
<description>how to communicate in the switch process</description>
</property>
<property>
<name>dfs.ha.fencing.ssh.private-key-files</name>
<value>/home/hadoop/.ssh/id_rsa</value>
<description>the location stored ssh key</description>
</property>
<property>
<name>dfs.ha.fencing.ssh.connect-timeout</name>
<value>1000</value>
</property>
<property>
<name>dfs.namenode.handler.count</name>
<value>8</value>
</property>
<property>
<name>dfs.journalnode.edits.dir</name>
<value>/home/hadoop/hadoop-2.4.0/journal/</value>
</property>
<property>
<name>dfs.client.failover.proxy.provider.myCluster</name>
<value>org.apache.hadoop.hdfs.server.namenode.ha.ConfiguredFailoverProxyProvider
</value>
</property>
三、配置文件同步到其他集群中。
四、任意ZK节点上执行:
hdfs zkfc -formatZK
五、启动ZKFC (ZookeeperFailoverController是用来监控NN状态,协助实现主备NN切换的,所以仅仅在主备NN节点上启动就行)
hadoop-daemon.sh start zkfc
启动后我们可以看到ZKFC进程:[hadoop@trainingHadoop01 ~]$ jps
3084 QuorumPeerMain
3292 Jps
3247 DFSZKFailoverController
六、启动用于主备NN之间同步元数据信息的共享存储系统JournalNode。在集群中各个节点上执行。
hadoop-daemon.sh start journalnode
[hadoop@trainingHadoop01 ~]$ jps
3084 QuorumPeerMain
3358 Jps
3325 JournalNode
3247 DFSZKFailoverController
七、格式化并启动主NN
hdfs namenode -format
注意:只有第一次启动系统时需格式化,请勿重复格式化!
在主NN节点执行命令启动NN:
hadoop-daemon.sh start namenode
启动后可以看到NN进程:
[hadoop@trainingHadoop01 ~]$ jps
3084 QuorumPeerMain
3480 Jps
3325 JournalNode
3411 NameNode
3247 DFSZKFailoverController
八、在备NN上同步主NN的元数据信息
hdfs namenode -bootstrapStandby
九、启动备NN
在备NN上执行命令:
hadoop-daemon.sh start namenode
十、设置主NN(这一步可以省略,这是在设置手动切换NN时的步骤,ZK已经自动选择一个节点作为主NN了)
到目前为止,其实HDFS还不知道谁是主NN,可以通过监控页面查看,两个节点的NN都是Standby状态。
下面我们需要在主NN节点上执行命令激活主NN:
hdfs haadmin -transitionToActive nn1
十一、主NN上启动Datanode
在[nn1]上,启动所有datanode
hadoop-daemons.sh start datanode
所有节点启动完成后,验证主备切换。
首先打开地址:
http://10.21.16.48:50070/dfshealth.html#tab-overview
Overview 'trainingHadoop02:8020' (standby)
http://10.21.16.47:50070/dfshealth.html#tab-overview
Overview 'trainingHadoop01:8020' (active)
这个时候kill 掉trainingHadoop01 机器上的namenode进程。
刷新http://10.21.16.48:50070/dfshealth.html#tab-overview 这个地址。展示的是
Overview 'trainingHadoop02:8020' (active)
可能出现的问题:
Standby NN continuing as standby when active NN machine got shutdown.---这是个bug.在hadoop-2.4.0上,不过可以通过配置来解决这个问题。
<property>
<name>dfs.ha.fencing.methods</name>
<value>shell(/bin/true)</value>
</property>
Yarn的ResourceManage的HA配置
主要是修改yarn-site.xml
<property>
<name>yarn.resourcemanager.ha.enabled</name>
<value>true</value>
</property>
<property>
<description>Enable automatic failover.
By default, it is enabled only when HA is enabled
</description>
<name>yarn.resourcemanager.ha.automatic-failover.enabled</name>
<value>true</value>
</property>
<property>
<name>yarn.resourcemanager.ha.rm-ids</name>
<value>rm1,rm2</value>
</property>
<property>
<name>yarn.resourcemanager.zk-address</name>
<value>trainingHadoop01:2181,trainingHadoop02:2181,trainingHadoop03:2181
</value>
</property>
<property>
<name>yarn.resourcemanager.hostname.rm1</name>
<value>trainingHadoop01</value>
</property>
<property>
<name>yarn.resourcemanager.hostname.rm2</name>
<value>trainingHadoop02</value>
</property>
<property>
<name>yarn.resourcemanager.ha.id</name>
<value>rm1</value>
<description>If we want to launch more than one RM in single node, we
need this configuration</description>
</property>
<!开启自动恢复功能>
<property>
<name>yarn.resourcemanager.recovery.enabled</name>
<value>true</value>
</property>
<property>
<name>yarn.resourcemanager.store.class</name>
<value>org.apache.hadoop.yarn.server.resourcemanager.recovery.ZKRMStateStore
</value>
</property>
<property>
<name>yarn.resourcemanager.cluster-id</name>
<value>yarn-cluster</value>
</property>
将修改后的配置文件yarn-site.xml同步到集群中另外一台机器上。
启动,yarn-daemon.sh start resourcemanager
访问http://trainingHadoop01:8088/
展示的是active的页面
在备用机器上也启动resourcemanager.
访问http://trainingHadoop01:8088/
展示的是standby的页面This is standby RM. Redirecting to the current active RM: http://trainingHadoop02:8088/cluster
然后会自动跳转到active的地址。
自动切换验证:
将active的机器的resourcemanager的进程kill掉。
standby的备用的机器变为active。