hadoop2.6.5 ha配置与yarn ha配置

小鱼儿 2022-06-11 05:10 162阅读 0赞

前提：java，ssh，hosts都配置完了。

master:  namenode  resourcemanager  zookeeper  zkfc

slave1:  datanode  journalnode  nodemanager   zookeeper

slave2:  datanode  journalnode  nodemanager   zookeeper

node1:  zkfc  namenode  resourcemanager

1.hdfs-site.xml

<property> <name>dfs.nameservices</name> <value>myha</value> </property> <property> <name>dfs.ha.namenodes.myha</name> <value>nn1,nn2</value> </property> <property> <name>dfs.namenode.rpc-address.myha.nn1</name> <value>master:8020</value> </property> <property> <name>dfs.namenode.rpc-address.myha.nn2</name> <value>node1:8020</value> </property> <property> <name>dfs.namenode.http-address.myha.nn1</name> <value>master:50070</value> </property> <property> <name>dfs.namenode.http-address.myha.nn2</name> <value>node1:50070</value> </property> <property> <name>dfs.namenode.shared.edits.dir</name> <value>qjournal://node1:8485;slave1:8485;slave2:8485/myha</value> </property> <property> <name>dfs.client.failover.proxy.provider.myha</name> <value>org.apache.hadoop.hdfs.server.namenode.ha.ConfiguredFailoverProxyProvider</value> </property> <property> <name>dfs.ha.fencing.methods</name> <value>shell(/bin/true)</value> </property> <property> <name>dfs.ha.fencing.ssh.private-key-files</name> <value>/home/hadoop/.ssh/id_rsa</value> </property> <property> <name>dfs.journalnode.edits.dir</name> <value>/usr/hadoop/dfs/journalnode</value> </property> <property> <name>dfs.ha.automatic-failover.enabled</name> <value>true</value> </property> <property> <name>dfs.replication</name> <value>2</value> </property> <property> <name>dfs.permissions.enabled</name> <value>false</value> </property>

2.core-site.xml

<property> <name>fs.defaultFS</name> <value>hdfs://myha</value> </property> <property> <name>ha.zookeeper.quorum</name> <value>master:2181,slave1:2181,slave2:2181</value> </property> <property> <name>hadoop.tmp.dir</name> <value>/usr/hadoop/tmp</value> </property>

3.yarn-site.xml

<property> <name>yarn.resourcemanager.ha.enabled</name> <value>true</value> </property> <property> <name>yarn.resourcemanager.cluster-id</name> <value>myyarn</value> </property> <property> <name>yarn.resourcemanager.ha.rm-ids</name> <value>rm1,rm2</value> </property> <property> <name>yarn.resourcemanager.hostname.rm1</name> <value>master</value> </property> <property> <name>yarn.resourcemanager.hostname.rm2</name> <value>node1</value> </property> <property> <name>yarn.resourcemanager.zk-address</name> <value>master:2181,slave1:2181,slave2:2181</value> </property> <property> <name>yarn.nodemanager.aux-services</name> <value>mapreduce_shuffle</value> </property> <property> <name>yarn.resourcemanager.ha.automatic-failover.enabled</name> <value>true</value> </property> <property> <name>yarn.resourcemanager.ha.automatic-failover.embedded</name> <value>true</value> </property> <property> <name>yarn.resourcemanager.ha.id</name> <value>rm1</value>  </property>

4.mapred-site.xml

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

5.运行hdfs

1.   先开启 zookeeper

2.   在开启journalnode      hadoop-daemon.sh start  journalnode

3.    namenode格式化       hdfs namenode -format

4.     zkfc格式化                 hdfs zkfc -formatZK

5.     master开启namenode       hadoop-daemon.sh start namenode

6.     master上开启datanode      hadoop-daemons.sh  start datanode

7.     master,node1上开启zkfc    hadoop-daemon.sh start zkfc

8.     node1备份                          hdfs namenode -bootstrapStandby

9.     node1开启namenode         hadoop-daemon.sh start namenode

上面命令以后可以用start-dfs.sh来代替。

验证：

[hadoop@master tmp]$ hdfs haadmin -getServiceState nn1 active [hadoop@master tmp]$ hdfs haadmin -getServiceState nn2 standby

master上kill掉namenode，看看会不会把standby的namenode变为active。

[hadoop@master tmp]$ jps 4833 Jps 4199 ResourceManager 3304 NameNode 2377 QuorumPeerMain 3581 DFSZKFailoverController [hadoop@master tmp]$ kill -9 3304 [hadoop@master tmp]$ hdfs haadmin -getServiceState nn2 active [hadoop@master tmp]$ hdfs haadmin -getServiceState nn1 17/08/01 20:30:57 INFO ipc.Client: Retrying connect to server: master/192.168.0.110:8020. Already tried 0 time(s); retry policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=1, sleepTime=1000 MILLISECONDS) Operation failed: Call From master/192.168.0.110 to master:8020 failed on connection exception: java.net.ConnectException: 拒绝连接; For more details see: http://wiki.apache.org/hadoop/ConnectionRefused [hadoop@master tmp]$

从上面可以看到是可以的。还可以用hdfs haadmin命令来转换

[hadoop@master tmp]$ hadoop-daemon.sh start namenode starting namenode, logging to /usr/hadoop/logs/hadoop-hadoop-namenode-master.out [hadoop@master tmp]$ jps 4199 ResourceManager 2377 QuorumPeerMain 5002 Jps 3581 DFSZKFailoverController 4926 NameNode [hadoop@master tmp]$ hdfs haadmin -getServiceState nn1 standby [hadoop@master tmp]$ hdfs haadmin -getServiceState nn2 active [hadoop@master tmp]$ hdfs haadmin -failover --forcefence --forceactive nn2 nn1 forcefence and forceactive flags not supported with auto-failover enabled. [hadoop@master tmp]$ hdfs haadmin -getServiceState nn2 active [hadoop@master tmp]$ hdfs haadmin -failover nn2 nn1 Failover to NameNode at master/192.168.0.110:8020 successful [hadoop@master tmp]$ hdfs haadmin -getServiceState nn2 standby [hadoop@master tmp]$ hdfs haadmin -getServiceState nn1 active

6.运行yarn

1.  master上运行resourcemanager     yarn-daemon.sh start resourcemanager

2.  node1上运行resourcemanager      yarn-daemon.sh start resourcemanager

3.  master上运行nodemanager           yarn-daemons.sh start nodemanager

验证：

[hadoop@master tmp]$ yarn rmadmin -getServiceState rm1 active [hadoop@master tmp]$ yarn rmadmin -getServiceState rm2 standby [hadoop@master tmp]$ jps 5728 Jps 5399 ResourceManager 2377 QuorumPeerMain 3581 DFSZKFailoverController 4926 NameNode [hadoop@master tmp]$ kill -9 5399 [hadoop@master tmp]$ yarn rmadmin -getServiceState rm2 standby [hadoop@master tmp]$ yarn rmadmin -getServiceState rm1 17/08/01 20:40:01 INFO ipc.Client: Retrying connect to server: master/192.168.0.110:8033. Already tried 0 time(s); retry policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=1, sleepTime=1000 MILLISECONDS) Operation failed: Call From master/192.168.0.110 to master:8033 failed on connection exception: java.net.ConnectException: 拒绝连接; For more details see: http://wiki.apache.org/hadoop/ConnectionRefused [hadoop@master tmp]$ yarn rmadmin -getServiceState rm2 active