Linux -- 之HDFS实现自动切换HA(全新HDFS)
JDK规划
1.7及以上 https://blog.****.net/meiLin_Ya/article/details/80650945
防火墙规划
系统防火墙关闭
SSH免密码规划
hadoop01(nn1)--> hadoop01(nn1) 需要免密码
hadoop01(nn1)--> hadoop02(nn2) 需要免密码
hadoop01(nn1)--> hadoop03(dn) 需要免密码
hadoop02(nn2)--> hadoop01(nn1) 需要免密码
hadoop02(nn2)--> hadoop02(nn2) 需要免密码
hadoop02(nn2)--> hadoop03(dn) 需要免密码
如果多节点之间全部免密码更好(生产环境不建议) 默认环境
Zk集群规划
已有可用zk集群 https://blog.****.net/meiLin_Ya/article/details/80654268
开始配置
首先我们要将所有的hadoop删除干净。如/temp /hadoopdata 等等,然后将hadoop的压缩包导入。你的集群中的每个节点也是,都要删除的。
删除后解压hadoop:
tar zxvf hadoop-2.6.0.tar.gz
修改core-site.xml
<configuration>
<property>
<name>fs.defaultFS</name>
<value>hdfs://beiwang</value>
</property>
<property>
<name>hadoop.tmp.dir</name>
<value>/home/hadoopdata/tmp</value>
</property>
<property>
<name>io.file.buffer.size</name>
<value>4096</value>
</property>
</configuration>
修改hdfs-site.xml
注意:中文注释不要带
<configuration> <!-- 指定hdfs的nameservice为beiwang,就是那个代理程序,询问zk集群哪个namenode还活着 --> <property> <name>dfs.nameservices</name> <value>beiwang</value> </property> <!—指定hdfs的两个NameNode都是什么名字(等会儿下面会配置他们所对应的机器的信息)--> <property> <name>dfs.ha.namenodes.beiwang</name> <value>nn1,nn2</value> </property> <!—NameNode1的rpc通讯地址--> <property> <name>dfs.namenode.rpc-address.beiwang.nn1</name> <value> hadoop01:8020</value> </property> <!—NameNode2的rpc通讯地址--> <property> <name>dfs.namenode.rpc-address.beiwang.nn2</name> <value> hadoop02:8020</value> </property> <!—NameNode1的web界面地址--> <property> <name>dfs.namenode.http-address.beiwang.nn1</name> <value> hadoop01:50070</value> </property> <!—NameNode2的web界面地址--> <property> <name>dfs.namenode.http-address.beiwang.nn2</name> <value> hadoop02:50070</value> </property> ######如果给一个有数据的HDFS添加HA,此处无需更改,保持原有地址即可##### <!---namenode存放元数据信息的Linux本地地址,这个目录不需要我们自己创建-> <property> <name>dfs.namenode.name.dir</name> <value>file:///home/hdfs/name</value> </property> <!—datanode存放用户提交的大文件的本地Linux地址,这个目录不需要我们自己创建--> <property> <name>dfs.datanode.data.dir</name> <value>file:///home/hdfs/data</value> </property> ########################################################### <!—QJM存放共享数据的方式--> <property> <name>dfs.namenode.shared.edits.dir</name> <value>qjournal:// hadoop01:8485; hadoop02:8485; hadoop03:8485/beiwang</value> </property> <!—单个QJM进程(角色)存放本地edits文件的Linux地址--> <property> <name>dfs.journalnode.edits.dir</name> <value>/home/bigdata/hadoop/journal1</value> </property> <!—开启hdfs的namenode死亡后自动切换--> <property> <name>dfs.ha.automatic-failover.enabled</name> <value>true</value> </property> <!-- 指定zookeeper集群地址,辅助两个namenode进行失败切换 --> <property> <name>ha.zookeeper.quorum</name> <value> hadoop01:2181, hadoop02:2181, hadoop03:2181</value> </property> <!—zkfc程序的主类--> <property> <name>dfs.client.failover.proxy.provider.beiwang</name> <value>org.apache.hadoop.hdfs.server.namenode.ha.ConfiguredFailoverProxyProvider</value> </property> <!—防止多个namenode同时active(脑裂)的方式--> <property> <name>dfs.ha.fencing.methods</name> <value>sshfence</value> </property> <!—指定本机的私钥所在目录--> <property> <name>dfs.ha.fencing.ssh.private-key-files</name> <value>/root/.ssh/id_rsa</value> </property> <!—指定ssh通讯超时时间--> <property> <name>dfs.ha.fencing.ssh.connect-timeout</name> <value>30000</value> </property> </configuration>
hadoop-env.sh
export JAVA_HOME="/home/bigdata/jdk1.8.0_161"
建一个master文本在hadoop-2.6.0/etc/hadoop/
注意:
新建master文件,该文件中写 所有namenode主机
hu-hadoop1
hu-hadoop2
hu-hadoop3
slaves:
hu-hadoop1
hu-hadoop2
hu-hadoop3
开启日志文件:
hadoop-daemons.sh start journalnode
启动zookeeper:
zkServer.sh start
然后进行格式化:
hadoop namenode -format
在master上开启namenode
hadoop-daemon.sh start namenode
在salve11机上 同步元数据信息
hdfs namenode -bootstrapStandby
格式化ZK(在Master上执行即可)
# hdfs zkfc -formatZK
格式化后可以查看zookeeper存放文件:
启动dfs:然后再查看zookeeper
start-dfs.sh
进入网页:
现在我们来测试下,杀死hu-hadoop2 ,然后看hu-hadoop1是否 可以从 standby=>active
然后我们再启动下刚刚的hu-hadoop2:查看它的状态
很明确的看出来了吧
然后我们看看zookeeper
然后现在我们去杀掉hu-hadoop1:再看zookeeper
当然此时网页hu-hadoop2又从刚刚的standby==>active
这是为什么呢?
记得我们前面做了一步 hdfs zkfc -formatZK 这一步就是将hdfs信息记录到zookeeper,还有hdfs-core.xml中的配置。
这就是zookeeper的强大之处。我们可以将zookeeper理解为数据库,而它和数据库又不太是因为他是一个树形,只有在树枝的末梢才会存储数据。它的大小有1MB,不要小看它的1MB,它的作用比你想象的要强大
再记录下我运行成功后的配置文件吧
hadoop-env.sh
export JAVA_HOME="/home/bigdata/jdk1.8.0_161"
core-site.xml
<configuration>
<property>
<name>fs.defaultFS</name>
<value>hdfs://huhu</value>
</property>
<property>
<name>hadoop.tmp.dir</name>
<value>/home/hadoopdata/tmp</value>
</property>
<property>
<name>io.file.buffer.size</name>
<value>4096</value>
</property>
</configuration>
hdfs-site.xml
<configuration>
<property>
<name>dfs.nameservices</name>
<value>huhu</value>
</property>
<property>
<name>dfs.ha.namenodes.huhu</name>
<value>huhu1,huhu2</value>
</property> <property>
<name>dfs.namenode.rpc-address.huhu.huhu1</name>
<value>hu-hadoop1:8020</value>
</property>
<property>
<name>dfs.namenode.rpc-address.huhu.huhu2</name>
<value>hu-hadoop2:8020</value>
</property>
<property>
<name>dfs.namenode.http-address.huhu.huhu1</name>
<value>hu-hadoop1:50070</value>
</property>
<property>
<name>dfs.namenode.http-address.huhu.huhu2</name>
<value>hu-hadoop2:50070</value>
</property>
<property>
<name>dfs.namenode.name.dir</name>
<value>file:///home/hdfs/name</value>
</property>
<property>
<name>dfs.datanode.data.dir</name>
<value>file:///home/hdfs/data</value>
</property>
<property>
<name>dfs.namenode.shared.edits.dir</name>
<value>qjournal://hu-hadoop1:8485;hu-hadoop2:8485;hu-hadoop3:8485/huhu</value>
</property>
<property>
<name>dfs.journalnode.edits.dir</name>
<value>/home/bigdata/hadoop/journal1</value>
</property>
<property>
<name>dfs.ha.automatic-failover.enabled</name>
<value>true</value>
</property>
<property>
<name>ha.zookeeper.quorum</name>
<value>hu-hadoop1:2181,hu-hadoop2:2181,hu-hadoop3:2181</value>
</property>
<property>
<name>dfs.client.failover.proxy.provider.huhu</name>
<value>org.apache.hadoop.hdfs.server.namenode.ha.ConfiguredFailoverProxyProvider</value>
</property>
<property>
<name>dfs.ha.fencing.methods</name>
<value>sshfence</value>
</property>
<property>
<name>dfs.ha.fencing.ssh.private-key-files</name>
<value>/root/.ssh/id_rsa</value>
</property>
<property>
<name>dfs.ha.fencing.ssh.connect-timeout</name>
<value>30000</value>
</property>
</configuration>
mapred-site.xml
<configuration>
<property>
<name>mapreduce.framework.name</name>
<value>yarn</value>
<final>true</final>
</property>
</configuration>
yarn-site.xml
<configuration>
<property>
<name>yarn.resourcemanager.connect.retry-interval.ms</name>
<value>2000</value>
</property>
<property>
<name>yarn.resourcemanager.ha.enabled</name>
<value>true</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.cluster-id</name>
<value>beiwangyarn</value>
</property>
<property>
<name>yarn.resourcemanager.ha.rm-ids</name>
<value>rm1,rm2</value>
</property>
<property>
<name>yarn.resourcemanager.hostname.rm1</name>
<value>hu-hadoop1</value>
</property>
<property>
<name>yarn.resourcemanager.hostname.rm2</name>
<value>hu-hadoop2</value>
</property>
<property>
<name>yarn.resourcemanager.scheduler.class</name>
<value>org.apache.hadoop.yarn.server.resourcemanager.scheduler.fair.FairScheduler</value>
</property>
<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.zk.state-store.address</name>
<value>hu-hadoop1:2181,hu-hadoop2:2181,hu-hadoop3:2181</value>
</property>
<property>
<name>yarn.app.mapreduce.am.scheduler.connection.wait.interval-ms</name>
<value>5000</value>
</property>
<property>
<name>yarn.resourcemanager.address.rm1</name>
<value>hu-hadoop1:8032</value>
</property>
<property>
<name>yarn.resourcemanager.scheduler.address.rm1</name>
<value>hu-hadoop1:8030</value>
</property>
<property>
<name>yarn.resourcemanager.webapp.https.address.rm1</name>
<value>hu-hadoop1:23189</value>
</property>
<property>
<name>yarn.resourcemanager.webapp.address.rm1</name>
<value>hu-hadoop1:8088</value>
</property>
<property>
<name>yarn.resourcemanager.resource-tracker.address.rm1</name>
<value>hu-hadoop1:8031</value>
</property>
<property>
<name>yarn.resourcemanager.admin.address.rm1</name>
<value>hu-hadoop1:8033</value>
</property>
<property>
<name>yarn.resourcemanager.address.rm2</name>
<value>hu-hadoop2:8032</value>
</property>
<property>
<name>yarn.resourcemanager.scheduler.address.rm2</name>
<value>hu-hadoop2:8030</value>
</property>
<property>
<name>yarn.resourcemanager.webapp.https.address.rm2</name>
<value>hu-hadoop2:23189</value>
</property>
<property>
<name>yarn.resourcemanager.webapp.address.rm2</name>
<value>hu-hadoop2:8088</value>
</property>
<property>
<name>yarn.resourcemanager.resource-tracker.address.rm2</name>
<value>hu-hadoop2:8031</value>
</property>
<property>
<name>yarn.resourcemanager.admin.address.rm2</name>
<value>hu-hadoop2:8033</value>
</property>
<property>
<description>Address where the localizer IPC is.</description>
<name>yarn.nodemanager.localizer.address</name>
<value>0.0.0.0:23344</value>
</property>
<property>
<description>NM Webapp address.</description>
<name>yarn.nodemanager.webapp.address</name>
<value>0.0.0.0:23999</value>
</property>
<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>
<property>
<name>yarn.nodemanager.local-dirs</name>
<value>/tmp/pseudo-dist/yarn/local</value>
</property>
<name>yarn.nodemanager.log-dirs</name>
<value>/tmp/pseudo-dist/yarn/log</value>
</property>
<property>
<name>mapreduce.shuffle.port</name>
<value>23080</value>
</property>
<property>
<name>yarn.resourcemanager.zk-address</name>
<value>hu-hadoop1:2181,hu-hadoop2:2181,hu-hadoop3:2181</value>
</property>
</configuration>
master
hu-hadoop1
hu-hadoop2
hu-hadoop3
slaves
hu-hadoop1
hu-hadoop2
hu-hadoop3