搭建hadoop HA高可用 原

Step1:集羣規劃

Step2:設置hosts

Step3:關閉防火牆

Step4:關閉Selinux

Step5:免密鑰登陸

Step6:安裝jdk

#解壓

tar -xvf jdk-8u131-linux-x64.tar.gz


mv jdk1.8.0_131 /usr/local/jdk1.8


#設置環境變量
vim /etc/profile

JAVA_HOME=/usr/local/jdk1.8/
JAVA_BIN=/usr/local/jdk1.8/bin
JRE_HOME=/usr/local/jdk1.8/jre
PATH=$PATH:/usr/local/jdk1.8/bin:/usr/local/jdk1.8/jre/bin
CLASSPATH=/usr/local/jdk1.8/jre/lib:/usr/local/jdk1.8/lib:/usr/local/jdk1.8/jre/lib/charsets.jar

Step7:安裝zookeeper

#下載 
[root@sl-opencron src]# wget http://www-eu.apache.org/dist/zookeeper/zookeeper-3.4.10/zookeeper-3.4.10.tar.gz

#解壓
[root@sl-opencron src]# tar -xvf zookeeper-3.4.10.tar.gz

#解壓後的目錄mv 到/usr/local/下
[root@sl-opencron src]# mv zookeeper-3.4.10 /usr/local/zookeeper

Step7.1:配置zookeeper

cd /usr/local/zookeeper/conf/

#將zoo_sample.cfg（樣板配置文件）命名爲zoo.cfg
mv zoo_sample.cfg  zoo.cfg

#修改配置文件
[root@sl-opencron conf]# vim zoo.cfg 

***********
***********

#路徑可自定義
dataDir=/data/zookeeper  


server.1=139.162.53.29:2888:3888
server.2=139.162.53.146:2888:3888
server.3=139.162.8.37:2888:3888

Step7.2:生成myid文件

mkdir /data/zookeeper

cd /data/zookeeper

touch myid

echo "1" >> myid

說明：cdh-1 myid是1  cdh-2 myid是2  cdh-3 myid是3

Step7.3:啓動zookeeper集羣

說明：分別在cdh-1 cdh-2 cdh-3 

cd /usr/local/zookeeper/bin

./zkServer.sh start

Step8:安裝hadoop

#下載
wget http://www-eu.apache.org/dist/hadoop/common/hadoop-2.7.6/hadoop-2.7.6.tar.gz

#解壓
tar -xvf hadoop-2.7.6.tar.gz

#解壓後的目錄移動到/usr/local/
 mv hadoop-2.7.6 /usr/local/hadoop

#進hadoop目錄
 cd /usr/local/hadoop

#創建幾個目錄
[root@hadooop-master hadoop]# mkdir tmp dfs dfs/data dfs/name

Step8.1:配置hadoop

vim core-site.xml


<configuration>
   <!-- 指定hdfs的nameservice爲ns1 -->
<property>
<name>fs.defaultFS</name>
<value>hdfs://ns1/</value>
</property>

<!-- 指定hadoop暫時文件夾 -->
<property>
<name>hadoop.tmp.dir</name>
<value>/usr/local/hadoop/tmp</value>
</property>

<!--指定zookeeper地址-->
<property>
<name>ha.zookeeper.quorum</name>
<value>cdh-1:2181,cdh-2:2181,cdh-3:2181</value>
</property>
</configuration>

 

vim hdfs-site.xml 

<configuration>	
<!--指定hdfs的nameservice爲ns1，須要和core-site.xml中的保持一致 -->
	<property>
		<name>dfs.nameservices</name>
		<value>ns1</value>
	</property>
	<!-- ns1以下有兩個NameNode，各自是nn1，nn2 -->
	<property>
		<name>dfs.ha.namenodes.ns1</name>
		<value>nn1,nn2</value>
	</property>
	<!-- nn1的RPC通信地址 -->
	<property>
		<name>dfs.namenode.rpc-address.ns1.nn1</name>
		<value>cdh-4:9000</value>
	</property>
	<!-- nn1的http通信地址 -->
	<property>
		<name>dfs.namenode.http-address.ns1.nn1</name>
		<value>cdh-4:50070</value>
	</property>
	<!-- nn2的RPC通信地址 -->
	<property>
		<name>dfs.namenode.rpc-address.ns1.nn2</name>
		<value>cdh-5:9000</value>
	</property>
	<!-- nn2的http通信地址 -->
	<property>
		<name>dfs.namenode.http-address.ns1.nn2</name>
		<value>cdh-5:50070</value>
	</property>
	<!-- 指定NameNode的元數據在JournalNode上的存放位置 -->
	<property>
		<name>dfs.namenode.shared.edits.dir</name>
		<value>qjournal://cdh-1:8485;cdh-2:8485;cdh-3:8485/ns1</value>
	</property>
	<!-- 指定JournalNode在本地磁盤存放數據的位置 -->
	<property>
		<name>dfs.journalnode.edits.dir</name>
		<value>/usr/local/hadoop/dfs/data</value>
	</property>
	<!-- 開啓NameNode失敗自己主動切換 -->
	<property>
		<name>dfs.ha.automatic-failover.enabled</name>
		<value>true</value>
	</property>
	<!-- 配置失敗自己主動切換實現方式 -->
	<property>
		<name>dfs.client.failover.proxy.provider.ns1</name>
		<value>org.apache.hadoop.hdfs.server.namenode.ha.ConfiguredFailoverProxyProvider</value>
	</property>
	<!-- 配置隔離機制方法。多個機制用換行切割，即每一個機制暫用一行-->
	<property>
		<name>dfs.ha.fencing.methods</name>
		<value>
			sshfence
			shell(/bin/true)
		</value>
	</property>
	<!-- 使用sshfence隔離機制時須要ssh免登陸 -->
	<property>
		<name>dfs.ha.fencing.ssh.private-key-files</name>
		<value>/root/.ssh/id_rsa</value>
	</property>
	<!-- 配置sshfence隔離機制超時時間 -->
	<property>
		<name>dfs.ha.fencing.ssh.connect-timeout</name>
		<value>30000</value>
	</property>
         
         <property>
         <name>ipc.client.connect.max.retries</name>
         <value>30</value>
  <property>
</configuration>

 

vim yarn.site.xml

<configuration>
<!-- Site specific YARN configuration properties -->
<!-- 開啓RM高可用 -->
	<property>
	   <name>yarn.resourcemanager.ha.enabled</name>
	   <value>true</value>
	</property>
	<!-- 指定RM的cluster id -->
	<property>
	   <name>yarn.resourcemanager.cluster-id</name>
	   <value>yarn-cluster-rm</value>
	</property>
	<!-- 指定RM的名字 -->
	<property>
	   <name>yarn.resourcemanager.ha.rm-ids</name>
	   <value>rm1,rm2</value>
	</property>
	<!-- 分別指定RM的地址 -->
	<property>
	   <name>yarn.resourcemanager.hostname.rm1</name>
	   <value>cdh-4</value>
	</property>
	<property>
	   <name>yarn.resourcemanager.hostname.rm2</name>
	   <value>cdh-5</value>
	</property>
	<!-- 指定zookeeper集羣地址 -->
	<property>
	   <name>yarn.resourcemanager.zk-address</name>
	   <value>cdh-1:2181,cdh-2:2181,cdh-3:2181</value>
	</property>
	<property>
	   <name>yarn.nodemanager.aux-services</name>
	   <value>mapreduce_shuffle</value>
	</property>
</configuration>

 

vim mapred-site.xml

<configuration>
<!-- 指定mr框架爲yarn方式 -->
	<property>
		<name>mapreduce.framework.name</name>
		<value>yarn</value>
	</property>
</configuration>

 

vim slaves 

cdh-1
cdh-2
cdh-3

Step8.2:同步到每臺服務器

rsync -av /usr/local/hadoop/etc/ cdh-1:/usr/local/hadoop/etc/

rsync -av /usr/local/hadoop/etc/ cdh-2:/usr/local/hadoop/etc/

rsync -av /usr/local/hadoop/etc/ cdh-3:/usr/local/hadoop/etc/

rsync -av /usr/local/hadoop/etc/ cdh-4:/usr/local/hadoop/etc/

rsync -av /usr/local/hadoop/etc/ cdh-5:/usr/local/hadoop/etc/

Step9：啓動journalnode

說明：分別在cdh-1  cdh-2  cdh-3啓動

cd /usr/local/hadoop/sbin/

./hadoop-daemon.sh start journalnode

Step10:格式化HDFS 

說明：在cdh-4操作

cd /usr/local/hadoop/bin/

./hdfs namenode -format


#注意：格式化之後須要把tmp文件夾拷給cdh-5（不然cdh-5的namenode起不來）

 cd /usr/local/hadoop/

 scp -r tmp/ cdh-5:/usr/local/hadoop/
VERSION                                                                                             100%  207   222.7KB/s   00:00    
fsimage_0000000000000000000.md5                                                                     100%   62    11.3KB/s   00:00    
fsimage_0000000000000000000                                                                         100%  321   327.3KB/s   00:00    
seen_txid                                                                                           100%    2     1.4KB/s   00:00    

Step11:格式化ZKFC

說明：在cdh-4操作

cd /usr/local/hadoop/bin/

./hdfs zkfc -formatZK

 

Step12:啓動yarn

說明：在cdh-4操作

cd /usr/local/hadoop/sbin/

 ./start-yarn.sh

 

Step13:cdh-5的resourcemanager須要手動單獨啓動

cd /usr/local/hadoop/sbin/

./yarn-daemon.sh start resourcemanager

Step14:查看集羣進程

[root@cdh-1 ~]# jps
26754 QuorumPeerMain
22387 JournalNode
5286 Jps
4824 NodeManager
25752 DataNode

[root@cdh-2 ~]# jps
4640 JournalNode
29520 QuorumPeerMain
5799 Jps
4839 DataNode
5642 NodeManager

[root@cdh-3 ~]# jps
28738 JournalNode
28898 DataNode
29363 NodeManager
20836 QuorumPeerMain
29515 Jps


[root@cdh-4 ~]# jps
21491 Jps
21334 NameNode
20167 DFSZKFailoverController
21033 ResourceManager

[root@cdh-5 ~]# jps
20403 ResourceManager
20280 NameNode
20523 Jps
19693 DFSZKFailoverController

Step15:測試高可用集羣

說明：如圖可見cdh-5的狀態是active  cdh-4的狀態是standby

Step15.1: 把cdh-5 Namenode stop 掉

[root@cdh-5 hadoop]# cd /usr/local/hadoop/sbin/

[root@cdh-5 sbin]# ./hadoop-daemon.sh stop namenode
stopping namenode


[root@cdh-5 sbin]# ./hadoop-daemon.sh start namenode

說明：將cdh-5 Namenode stop 掉，刷新cdh-4 可以看到現在cdh-4的狀態是active ，cdh-5 的狀態是standby