您的位置: 首页  >  文章  >  Hadoop-2.3.0-cdh5.0.1完全分布式环境搭建（NameNode,ResourceManager HA）

Hadoop-2.3.0-cdh5.0.1完全分布式环境搭建（NameNode,ResourceManager HA）

分类: 文章 • 2024-11-05 08:24:34

本文搭建Hadoop CDH5.0.1 分布式系统，包括NameNode ，ResourceManger HA，忽略了Web Application Proxy 和Job HistoryServer。 

        word版：见附件吧！

一概述

（一）HDFS

1）基础架构

（1）NameNode（Master）

  • 命名空间管理：命名空间支持对HDFS中的目录、文件和块做类似文件系统的创建、修改、删除、列表文件和目录等基本操作。
  • 块存储管理

（2）DataNode（Slaver）

namenode和client的指令进行存储或者检索block，并且周期性的向namenode节点报告它存了哪些文件的block

2）HA架构

 
Hadoop-2.3.0-cdh5.0.1完全分布式环境搭建（NameNode,ResourceManager HA）
 

使用Active NameNode，Standby NameNode 两个结点解决单点问题，两个结点通过JounalNode共享状态，通过ZKFC 选举Active ，监控状态，自动备援。

（1）Active NameNode：

接受client的RPC请求并处理，同时写自己的Editlog和共享存储上的Editlog，接收DataNode的Block report, block location updates和heartbeat；

（2）Standby NameNode：

      同样会接到来自DataNode的Block report, block location updates和heartbeat，同时会从共享存储的Editlog上读取并执行这些log操作，使得自己的NameNode中的元数据（Namespcae information + Block locations map）都是和Active NameNode中的元数据是同步的。所以说Standby模式的NameNode是一个热备（Hot Standby NameNode），一旦切换成Active模式，马上就可以提供NameNode服务

（3）JounalNode：

用于Active NameNode ， Standby NameNode 同步数据，本身由一组JounnalNode结点组成，该组结点基数个，支持Paxos协议，保证高可用，是CDH5唯一支持的共享方式（相对于CDH4 促在NFS共享方式）

（4）ZKFC：

监控NameNode进程，自动备援。

（二）YARN

1）基础架构

（1）ResourceManager(RM)

接收客户端任务请求，接收和监控NodeManager(NM)的资源情况汇报，负责资源的分配与调度，启动和监控ApplicationMaster(AM)。

（2）NodeManager

节点上的资源管理，启动Container运行task计算，上报资源、container情况给RM和任务处理情况给AM。

（3）ApplicationMaster

单个Application(Job)的task管理和调度，向RM进行资源的申请，向NM发出launch Container指令，接收NM的task处理状态信息。NodeManager

（4）Web Application Proxy

用于防止Yarn遭受Web攻击，本身是ResourceManager的一部分，可通过配置独立进程。ResourceManager Web的访问基于守信用户，当Application Master运行于一个非受信用户，其提供给ResourceManager的可能是非受信连接，Web Application Proxy可以阻止这种连接提供给RM。

（5）Job History Server

NodeManager在启动的时候会初始化LogAggregationService服务, 该服务会在把本机执行的container log (在container结束的时候)收集并存放到hdfs指定的目录下. ApplicationMaster会把jobhistory信息写到hdfs的jobhistory临时目录下, 并在结束的时候把jobhisoty移动到最终目录, 这样就同时支持了job的recovery.History会启动web和RPC服务, 用户可以通过网页或RPC方式获取作业的信息

2）HA架构

 
Hadoop-2.3.0-cdh5.0.1完全分布式环境搭建（NameNode,ResourceManager HA）
 

      ResourceManager HA 由一对Active，Standby结点构成，通过RMStateStore存储内部数据和主要应用的数据及标记。目前支持的可替代的RMStateStore实现有：基于内存的MemoryRMStateStore，基于文件系统的FileSystemRMStateStore，及基于zookeeper的ZKRMStateStore。

ResourceManager HA的架构模式同NameNode HA的架构模式基本一致，数据共享由RMStateStore，而ZKFC成为 ResourceManager进程的一个服务，非独立存在。

 二 规划

（一）版本

组件名

版本

说明

JRE

java version "1.7.0_60"

Java(TM) SE Runtime Environment (build 1.7.0_60-b19)

Java HotSpot(TM) Client VM (build 24.60-b09, mixed mode)

 

Hadoop

hadoop-2.3.0-cdh5.0.1.tar.gz

主程序包

 

Zookeeper

zookeeper-3.4.5-cdh5.0.1.tar.gz

热切,Yarn 存储数据使用的协调服务

（二）主机规划

IP

Host

部署模块

进程

8.8.8.11

Hadoop-NN-01

NameNode

ResourceManager

NameNode

DFSZKFailoverController

ResourceManager

8.8.8.12

Hadoop-NN-02

NameNode

ResourceManager

NameNode

DFSZKFailoverController

ResourceManager

8.8.8.13

Hadoop-DN-01

Zookeeper-01

DataNode

NodeManager

Zookeeper

DataNode

NodeManager

JournalNode

QuorumPeerMain

8.8.8.14

Hadoop-DN-02

Zookeeper-02

DataNode

NodeManager

Zookeeper

DataNode

NodeManager

JournalNode

QuorumPeerMain

8.8.8.15

Hadoop-DN-03

Zookeeper-03

DataNode

NodeManager

Zookeeper

DataNode

NodeManager

JournalNode

QuorumPeerMain

各个进程解释：

  • NameNode
  • ResourceManager
  • DFSZKFC：DFS Zookeeper Failover Controller **Standby NameNode
  • DataNode
  • NodeManager
  • JournalNode：NameNode共享editlog结点服务（如果使用NFS共享，则该进程和所有启动相关配置接可省略）。
  • QuorumPeerMain：Zookeeper主进程

（三）目录规划

名称

路径

$HADOOP_HOME

/home/zero/hadoop/hadoop-2.3.0-cdh5.0.1

Data

$ HADOOP_HOME/data

Log

$ HADOOP_HOME/logs

 三 环境准备

1）关闭防火墙

root 用户：

[[email protected] hadoop-2.3.0-cdh5.0.1]# service iptables stop

iptables: Flushing firewall rules: [  OK  ]

iptables: Setting chains to policy ACCEPT: filter [  OK  ]

iptables: Unloading modules: [  OK  ]

验证：

[[email protected] hadoop-2.3.0-cdh5.0.1]# service  iptables status

iptables: Firewall is not running.

2）安装JRE：略

3）安装Zookeeper ：参见《Zookeeper-3.4.5-cdh5.0.1 单机模式、副本模式安装、配置说明》

4）配置SSH互信：

（1）Hadoop-NN-01创建**：

[[email protected] ~]$ ssh-******

Generating public/private rsa key pair.

Enter file in which to save the key (/home/zero/.ssh/id_rsa):

Created directory '/home/zero/.ssh'.

Enter passphrase (empty for no passphrase):

Enter same passphrase again:

Your identification has been saved in /home/zero/.ssh/id_rsa.

Your public key has been saved in /home/zero/.ssh/id_rsa.pub.

The key fingerprint is:

28:0a:29:1d:98:56:55:db:ec:83:93:56:8a:0f:6c:ea [email protected]

The key's randomart image is:

+--[ RSA 2048]----+

|   .....         |

| o.     +        |

|o..    . +       |

|.o .. ..*        |

|+ . .=.*So       |

|.. .o.+ . .      |

|  ..   .         |

|  .              |

|   E             |

+-----------------+

（2）分发**：

[[email protected] ~]$ ssh-copy-id -i ~/.ssh/id_rsa.pub [email protected]

The authenticity of host 'hadoop-nn-01 (8.8.8.11)' can't be established.

RSA key fingerprint is a6:11:09:49:8c:fe:b2:fb:49:d5:01:fa:13:1b:32:24.

Are you sure you want to continue connecting (yes/no)? yes

Warning: Permanently added 'hadoop-nn-01,8.8.8.11' (RSA) to the list of known hosts.

[email protected]'s password:

Permission denied, please try again.

[email protected]'s password:

[[email protected] ~]$  ssh-copy-id -i ~/.ssh/id_rsa.pub [email protected]

[email protected]'s password:

Now try logging into the machine, with "ssh '[email protected]'", and check in:

 

  .ssh/authorized_keys

 

to make sure we haven't added extra keys that you weren't expecting.

[[email protected] ~]$  ssh-copy-id -i ~/.ssh/id_rsa.pub [email protected]

（…略…）

[[email protected] ~]$  ssh-copy-id -i ~/.ssh/id_rsa.pub [email protected]

（…略…）

[[email protected] ~]$  ssh-copy-id -i ~/.ssh/id_rsa.pub [email protected]

（…略…）

[[email protected] ~]$  ssh-copy-id -i ~/.ssh/id_rsa.pub [email protected]

（…略…）

（3）验证：

[[email protected] ~]$ ssh Hadoop-NN-01

Last login: Sun Jun 22 19:56:23 2014 from 8.8.8.1

[[email protected] ~]$ exit

logout

Connection to Hadoop-NN-01 closed.

[[email protected] ~]$ ssh Hadoop-NN-02

Last login: Sun Jun 22 20:03:31 2014 from 8.8.8.1

[[email protected] ~]$ exit

logout

Connection to Hadoop-NN-02 closed.

[[email protected] ~]$ ssh Hadoop-DN-01

Last login: Mon Jun 23 02:00:07 2014 from centos_cluster_01

[[email protected] ~]$ exit

logout

Connection to Hadoop-DN-01 closed.

[[email protected] ~]$ ssh Hadoop-DN-02

Last login: Sun Jun 22 20:07:03 2014 from 8.8.8.1

[[email protected] ~]$ exit

logout

Connection to Hadoop-DN-02 closed.

[[email protected] ~]$ ssh Hadoop-DN-03

Last login: Sun Jun 22 20:07:05 2014 from 8.8.8.1

[[email protected] ~]$ exit

logout

Connection to Hadoop-DN-03 closed.

5）配置/etc/hosts并分发：

[[email protected] zero]# vi /etc/hosts

127.0.0.1   localhost localhost.localdomain localhost4 localhost4.localdomain4

::1         localhost localhost.localdomain localhost6 localhost6.localdomain6

8.8.8.10 CentOS-StandAlone

8.8.8.11 CentOS-Cluster-01 Hadoop-NN-01

8.8.8.12 CentOS-Cluster-02 Hadoop-NN-02

8.8.8.13 CentOS-Cluster-03 Hadoop-DN-01 Zookeeper-01

8.8.8.14 CentOS-Cluster-04 Hadoop-DN-02 Zookeeper-02

8.8.8.15 CentOS-Cluster-05 Hadoop-DN-03 Zookeeper-03

6）配置环境变量：vi ~/.bashrc 然后 source ~/.bashrc

[[email protected] ~]$ vi ~/.bashrc

……

 

# hadoop cdh5

export HADOOP_HOME=/home/zero/hadoop/hadoop-2.3.0-cdh5.0.1

export PATH=$PATH:$HADOOP_HOME/sbin:$HADOOP_HOME/bin

[[email protected] ~]$ source ~/.bashrc

四 安装

1）解压

[[email protected] cdh4.4]$ tar -xvf hadoop-2.0.0-cdh4.4.0.tar.gz

2）修改配置文件

说明：

配置名称

类型

说明

hadoop-env.sh

Bash脚本

Hadoop运行环境变量设置

core-site.xml

xml

配置Hadoop core，如IO

hdfs-site.xml

xml

配置HDFS守护进程：NN、JN、DN

yarn-env.sh

Bash脚本

Yarn运行环境变量设置

yarn-site.xml

xml

Yarn框架配置环境

mapred-site.xml

xml

MR属性设置

capacity-scheduler.xml

xml

Yarn调度属性设置

container-executor.cfg

Cfg

Yarn Container配置

mapred-queues.xml

xml

MR队列设置

hadoop-metrics.properties

Java属性

Hadoop Metrics配置

hadoop-metrics2.properties

Java属性

Hadoop Metrics配置

slaves

Plain Text

DN节点配置

exclude

Plain Text

移除DN节点配置文件

log4j.properties

 

系统日志设置

configuration.xsl

 

 

（1）修改$HADOOP_HOME/etc/hadoop/hadoop-env.sh：

Shell代码  Hadoop-2.3.0-cdh5.0.1完全分布式环境搭建（NameNode,ResourceManager HA）
  1. #--------------------Java Env------------------------------  
  2. export JAVA_HOME="/usr/runtime/java/jdk1.7.0_60"  
  3.   
  4. #--------------------Hadoop Env------------------------------  
  5. #export HADOOP_PID_DIR=  
  6. export HADOOP_PREFIX="/home/zero/hadoop/hadoop-2.3.0-cdh5.0.1"  
  7.   
  8. #--------------------Hadoop Daemon Options-----------------  
  9. #export HADOOP_NAMENODE_OPTS="-XX:+UseParallelGC ${HADOOP_NAMENODE_OPTS}"  
  10. #export HADOOP_DATANODE_OPTS=  
  11.   
  12. #--------------------Hadoop Logs---------------------------  
  13. #export HADOOP_LOG_DIR=  

（2）修改$HADOOP_HOME/etc/hadoop/hadoop-site.xml

Xml代码  Hadoop-2.3.0-cdh5.0.1完全分布式环境搭建（NameNode,ResourceManager HA）
  1. <?xml version="1.0" encoding="UTF-8"?>  
  2. <?xml-stylesheet type="text/xsl" href="configuration.xsl"?>  
  3. <configuration>  
  4.     <!--Yarn 需要使用 fs.defaultFS 指定NameNode URI -->  
  5.     <property>  
  6.         <name>fs.defaultFS</name>  
  7.         <value>hdfs://mycluster</value>  
  8.     </property>  
  9.     <!--HDFS超级用户 -->  
  10.     <property>  
  11.         <name>dfs.permissions.superusergroup</name>  
  12.         <value>zero</value>  
  13.     </property>  
  14.     <!--==============================Trash机制======================================= -->  
  15.     <property>  
  16.         <!--多长时间创建CheckPoint NameNode截点上运行的CheckPointer 从Current文件夹创建CheckPoint;默认：0 由fs.trash.interval项指定 -->  
  17.         <name>fs.trash.checkpoint.interval</name>  
  18.         <value>0</value>  
  19.     </property>  
  20.     <property>  
  21.         <!--多少分钟.Trash下的CheckPoint目录会被删除,该配置服务器设置优先级大于客户端，默认：0 不删除 -->  
  22.         <name>fs.trash.interval</name>  
  23.         <value>1440</value>  
  24.     </property>  
  25. </configuration>  

 （3）修改$HADOOP_HOME/etc/hadoop/hdfs-site.xml

Xml代码  Hadoop-2.3.0-cdh5.0.1完全分布式环境搭建（NameNode,ResourceManager HA）
  1. <?xml version="1.0" encoding="UTF-8"?>  
  2. <?xml-stylesheet type="text/xsl" href="configuration.xsl"?>  
  3. <configuration>  
  4.     <!--开启web hdfs -->  
  5.     <property>  
  6.         <name>dfs.webhdfs.enabled</name>  
  7.         <value>true</value>  
  8.     </property>  
  9.     <property>  
  10.         <name>dfs.namenode.name.dir</name>  
  11.         <value>/home/zero/hadoop/hadoop-2.3.0-cdh5.0.1/data/dfs/name</value>  
  12.         <description> namenode 存放name table(fsimage)本地目录（需要修改）</description>  
  13.     </property>  
  14.     <property>  
  15.         <name>dfs.namenode.edits.dir</name>  
  16.         <value>${dfs.namenode.name.dir}</value>  
  17.         <description>namenode粗放 transaction file(edits)本地目录（需要修改）</description>  
  18.     </property>  
  19.     <property>  
  20.         <name>dfs.datanode.data.dir</name>  
  21.         <value>/home/zero/hadoop/hadoop-2.3.0-cdh5.0.1/data/dfs/data</value>  
  22.         <description>datanode存放block本地目录（需要修改）</description>  
  23.     </property>  
  24.     <property>  
  25.         <name>dfs.replication</name>  
  26.         <value>1</value>  
  27.     </property>  
  28.     <!-- 块大小 （默认） -->  
  29.     <property>  
  30.         <name>dfs.blocksize</name>  
  31.         <value>268435456</value>  
  32.     </property>  
  33.     <!--======================================================================= -->  
  34.     <!--HDFS高可用配置 -->  
  35.     <!--nameservices逻辑名 -->  
  36.     <property>  
  37.         <name>dfs.nameservices</name>  
  38.         <value>mycluster</value>  
  39.     </property>  
  40.     <property>  
  41.         <!--设置NameNode IDs 此版本最大只支持两个NameNode -->  
  42.         <name>dfs.ha.namenodes.mycluster</name>  
  43.         <value>nn1,nn2</value>  
  44.     </property>  
  45.     <!-- Hdfs HA: dfs.namenode.rpc-address.[nameservice ID] rpc 通信地址 -->  
  46.     <property>  
  47.         <name>dfs.namenode.rpc-address.mycluster.nn1</name>  
  48.         <value>Hadoop-NN-01:8020</value>  
  49.     </property>  
  50.     <property>  
  51.         <name>dfs.namenode.rpc-address.mycluster.nn2</name>  
  52.         <value>Hadoop-NN-02:8020</value>  
  53.     </property>  
  54.     <!-- Hdfs HA: dfs.namenode.http-address.[nameservice ID] http 通信地址 -->  
  55.     <property>  
  56.         <name>dfs.namenode.http-address.mycluster.nn1</name>  
  57.         <value>Hadoop-NN-01:50070</value>  
  58.     </property>  
  59.     <property>  
  60.         <name>dfs.namenode.http-address.mycluster.nn2</name>  
  61.         <value>Hadoop-NN-02:50070</value>  
  62.     </property>  
  63.   
  64.     <!--==================Namenode editlog同步 ============================================ -->  
  65.     <!--保证数据恢复 -->  
  66.     <property>  
  67.         <name>dfs.journalnode.http-address</name>  
  68.         <value>0.0.0.0:8480</value>  
  69.     </property>  
  70.     <property>  
  71.         <name>dfs.journalnode.rpc-address</name>  
  72.         <value>0.0.0.0:8485</value>  
  73.     </property>  
  74.     <property>  
  75.         <!--设置JournalNode服务器地址，QuorumJournalManager 用于存储editlog -->  
  76.         <!--格式：qjournal://<host1:port1>;<host2:port2>;<host3:port3>/<journalId> 端口同journalnode.rpc-address -->  
  77.         <name>dfs.namenode.shared.edits.dir</name>  
  78.         <value>qjournal://Hadoop-DN-01:8485;Hadoop-DN-02:8485;Hadoop-DN-03:8485/mycluster</value>  
  79.     </property>  
  80.     <property>  
  81.         <!--JournalNode存放数据地址 -->  
  82.         <name>dfs.journalnode.edits.dir</name>  
  83.         <value>/home/zero/hadoop/hadoop-2.3.0-cdh5.0.1/data/dfs/jn</value>  
  84.     </property>  
  85.     <!--==================DataNode editlog同步 ============================================ -->  
  86.     <property>  
  87.         <!--DataNode,Client连接Namenode识别选择Active NameNode策略 -->  
  88.         <name>dfs.client.failover.proxy.provider.mycluster</name>  
  89.         <value>org.apache.hadoop.hdfs.server.namenode.ha.ConfiguredFailoverProxyProvider</value>  
  90.     </property>  
  91.     <!--==================Namenode fencing：=============================================== -->  
  92.     <!--Failover后防止停掉的Namenode启动，造成两个服务 -->  
  93.     <property>  
  94.         <name>dfs.ha.fencing.methods</name>  
  95.         <value>sshfence</value>  
  96.     </property>  
  97.     <property>  
  98.         <name>dfs.ha.fencing.ssh.private-key-files</name>  
  99.         <value>/home/zero/.ssh/id_rsa</value>  
  100.     </property>  
  101.     <property>  
  102.         <!--多少milliseconds 认为fencing失败 -->  
  103.         <name>dfs.ha.fencing.ssh.connect-timeout</name>  
  104.         <value>30000</value>  
  105.     </property>  
  106.   
  107.     <!--==================NameNode auto failover base ZKFC and Zookeeper====================== -->  
  108.     <!--开启基于Zookeeper及ZKFC进程的自动备援设置,监视进程是否死掉 -->  
  109.     <property>  
  110.         <name>dfs.ha.automatic-failover.enabled</name>  
  111.         <value>true</value>  
  112.     </property>  
  113.     <property>  
  114.         <name>ha.zookeeper.quorum</name>  
  115.         <value>Zookeeper-01:2181,Zookeeper-02:2181,Zookeeper-03:2181</value>  
  116.     </property>  
  117.     <property>  
  118.         <!--指定ZooKeeper超时间隔，单位毫秒 -->  
  119.         <name>ha.zookeeper.session-timeout.ms</name>  
  120.         <value>2000</value>  
  121.     </property>  
  122. </configuration>  

 （4）修改$HADOOP_HOME/etc/hadoop/yarn-env.sh

Shell代码  Hadoop-2.3.0-cdh5.0.1完全分布式环境搭建（NameNode,ResourceManager HA）
  1. #Yarn Daemon Options  
  2. #export YARN_RESOURCEMANAGER_OPTS  
  3. #export YARN_NODEMANAGER_OPTS  
  4. #export YARN_PROXYSERVER_OPTS  
  5. #export HADOOP_JOB_HISTORYSERVER_OPTS  
  6.   
  7. #Yarn Logs  
  8. export YARN_LOG_DIR=” /home/zero/hadoop/hadoop-2.3.0-cdh5.0.1/logs”  

 （5）$HADOOP_HOEM/etc/hadoop/mapred-site.xml

Xml代码  Hadoop-2.3.0-cdh5.0.1完全分布式环境搭建（NameNode,ResourceManager HA）
  1. <configuration>  
  2.     <!-- 配置 MapReduce Applications -->  
  3.     <property>  
  4.         <name>mapreduce.framework.name</name>  
  5.         <value>yarn</value>  
  6.     </property>  
  7.     <!-- JobHistory Server ============================================================== -->  
  8.     <!-- 配置 MapReduce JobHistory Server 地址 ，默认端口10020 -->  
  9.     <property>  
  10.         <name>mapreduce.jobhistory.address</name>  
  11.         <value>0.0.0.0:10020</value>  
  12.     </property>  
  13.     <!-- 配置 MapReduce JobHistory Server web ui 地址， 默认端口19888 -->  
  14.     <property>  
  15.         <name>mapreduce.jobhistory.webapp.address</name>  
  16.         <value>0.0.0.0:19888</value>  
  17.     </property>  
  18. </configuration>  

 （6）修改$HADOOP_HOME/etc/hadoop/yarn-site.xml

Xml代码  Hadoop-2.3.0-cdh5.0.1完全分布式环境搭建（NameNode,ResourceManager HA）
  1. <configuration>  
  2.     <!-- nodemanager 配置 ================================================= -->  
  3.     <property>  
  4.         <name>yarn.nodemanager.aux-services</name>  
  5.         <value>mapreduce_shuffle</value>  
  6.     </property>  
  7.     <property>  
  8.         <name>yarn.nodemanager.aux-services.mapreduce.shuffle.class</name>  
  9.         <value>org.apache.hadoop.mapred.ShuffleHandler</value>  
  10.     </property>  
  11.     <property>  
  12.         <description>Address where the localizer IPC is.</description>  
  13.         <name>yarn.nodemanager.localizer.address</name>  
  14.         <value>0.0.0.0:23344</value>  
  15.     </property>  
  16.     <property>  
  17.         <description>NM Webapp address.</description>  
  18.         <name>yarn.nodemanager.webapp.address</name>  
  19.         <value>0.0.0.0:23999</value>  
  20.     </property>  
  21.   
  22.     <!-- HA 配置 =============================================================== -->  
  23.     <!-- Resource Manager Configs -->  
  24.     <property>  
  25.         <name>yarn.resourcemanager.connect.retry-interval.ms</name>  
  26.         <value>2000</value>  
  27.     </property>  
  28.     <property>  
  29.         <name>yarn.resourcemanager.ha.enabled</name>  
  30.         <value>true</value>  
  31.     </property>  
  32.     <property>  
  33.         <name>yarn.resourcemanager.ha.automatic-failover.enabled</name>  
  34.         <value>true</value>  
  35.     </property>  
  36.     <!-- 使嵌入式自动故障转移。HA环境启动，与 ZKRMStateStore 配合 处理fencing -->  
  37.     <property>  
  38.         <name>yarn.resourcemanager.ha.automatic-failover.embedded</name>  
  39.         <value>true</value>  
  40.     </property>  
  41.     <!-- 集群名称，确保HA选举时对应的集群 -->  
  42.     <property>  
  43.         <name>yarn.resourcemanager.cluster-id</name>  
  44.         <value>yarn-cluster</value>  
  45.     </property>  
  46.     <property>  
  47.         <name>yarn.resourcemanager.ha.rm-ids</name>  
  48.         <value>rm1,rm2</value>  
  49.     </property>  
  50.     <!—这里RM主备结点需要单独指定,（可选）  
  51.     <property>  
  52.         <name>yarn.resourcemanager.ha.id</name>  
  53.         <value>rm2</value>  
  54. </property>  
  55. -->  
  56.     <property>  
  57.         <name>yarn.resourcemanager.scheduler.class</name>  
  58.         <value>org.apache.hadoop.yarn.server.resourcemanager.scheduler.fair.FairScheduler</value>  
  59.     </property>  
  60.     <property>  
  61.         <name>yarn.resourcemanager.recovery.enabled</name>  
  62.         <value>true</value>  
  63.     </property>  
  64.     <property>  
  65.         <name>yarn.app.mapreduce.am.scheduler.connection.wait.interval-ms</name>  
  66.         <value>5000</value>  
  67.     </property>  
  68.     <!-- ZKRMStateStore 配置 -->  
  69.     <property>  
  70.         <name>yarn.resourcemanager.store.class</name>  
  71.         <value>org.apache.hadoop.yarn.server.resourcemanager.recovery.ZKRMStateStore</value>  
  72.     </property>  
  73.     <property>  
  74.         <name>yarn.resourcemanager.zk-address</name>  
  75.         <value>Zookeeper-01:2181,Zookeeper-02:2181,Zookeeper-03:2181</value>  
  76.     </property>  
  77.     <property>  
  78.         <name>yarn.resourcemanager.zk.state-store.address</name>  
  79.         <value>Zookeeper-01:2181,Zookeeper-02:2181,Zookeeper-03:2181</value>  
  80.     </property>  
  81.     <!-- Client访问RM的RPC地址 (applications manager interface) -->  
  82.     <property>  
  83.         <name>yarn.resourcemanager.address.rm1</name>  
  84.         <value>Hadoop-NN-01:23140</value>  
  85.     </property>  
  86.     <property>  
  87.         <name>yarn.resourcemanager.address.rm2</name>  
  88.         <value>Hadoop-NN-02:23140</value>  
  89.     </property>  
  90.     <!-- AM访问RM的RPC地址(scheduler interface) -->  
  91.     <property>  
  92.         <name>yarn.resourcemanager.scheduler.address.rm1</name>  
  93.         <value>Hadoop-NN-01:23130</value>  
  94.     </property>  
  95.     <property>  
  96.         <name>yarn.resourcemanager.scheduler.address.rm2</name>  
  97.         <value>Hadoop-NN-02:23130</value>  
  98.     </property>  
  99.     <!-- RM admin interface -->  
  100.     <property>  
  101.         <name>yarn.resourcemanager.admin.address.rm1</name>  
  102.         <value>Hadoop-NN-01:23141</value>  
  103.     </property>  
  104.     <property>  
  105.         <name>yarn.resourcemanager.admin.address.rm2</name>  
  106.         <value>Hadoop-NN-02:23141</value>  
  107.     </property>  
  108.     <!--NM访问RM的RPC端口 -->  
  109.     <property>  
  110.         <name>yarn.resourcemanager.resource-tracker.address.rm1</name>  
  111.         <value>Hadoop-NN-01:23125</value>  
  112.     </property>  
  113.     <property>  
  114.         <name>yarn.resourcemanager.resource-tracker.address.rm2</name>  
  115.         <value>Hadoop-NN-02:23125</value>  
  116.     </property>  
  117.     <!-- RM web application 地址 -->  
  118.     <property>  
  119.         <name>yarn.resourcemanager.webapp.address.rm1</name>  
  120.         <value>Hadoop-NN-01:23188</value>  
  121.     </property>  
  122.     <property>  
  123.         <name>yarn.resourcemanager.webapp.address.rm2</name>  
  124.         <value>Hadoop-NN-02:23188</value>  
  125.     </property>  
  126.     <property>  
  127.         <name>yarn.resourcemanager.webapp.https.address.rm1</name>  
  128.         <value>Hadoop-NN-01:23189</value>  
  129.     </property>  
  130.     <property>  
  131.         <name>yarn.resourcemanager.webapp.https.address.rm2</name>  
  132.         <value>Hadoop-NN-02:23189</value>  
  133.     </property>  
  134. </configuration>  

（7）修改slaves

[[email protected]  hadoop]$ vi slaves

Hadoop-DN-01

Hadoop-DN-02

Hadoop-DN-03

3）分发程序

[[email protected] ~]$ scp -r /home/zero/hadoop/hadoop-2.3.0-cdh5.0.1 [email protected]: /home/zero/hadoop/

…….

[[email protected] ~]$ scp -r /home/zero/hadoop/hadoop-2.3.0-cdh5.0.1 [email protected]: /home/zero/hadoop/

…….

[[email protected] ~]$ scp -r /home/zero/hadoop/hadoop-2.3.0-cdh5.0.1 [email protected]: /home/zero/hadoop/

…….

[[email protected] ~]$ scp -r /home/zero/hadoop/hadoop-2.3.0-cdh5.0.1 [email protected]: /home/zero/hadoop/

…….

4）启动HDFS

（1）启动JournalNode

格式化前需要在JournalNode结点上启动JournalNode：

[[email protected] hadoop-2.3.0-cdh5.0.1]$ hadoop-daemon.sh start journalnode

starting journalnode, logging to /home/puppet/hadoop/cdh4.4/hadoop-2.0.0-cdh4.4.0/logs/hadoop-puppet-journalnode-BigData-03.out

验证JournalNode：

[[email protected] hadoop-2.3.0-cdh5.0.1]$ jps

25918 QuorumPeerMain

16728 JournalNode

16772 Jps

（2）NameNode 格式化：

结点Hadoop-NN-01：hdfs namenode -format

[[email protected] hadoop-2.3.0-cdh5.0.1]$ hdfs namenode -format

14/06/23 21:02:49 INFO namenode.NameNode: STARTUP_MSG:

/************************************************************

STARTUP_MSG: Starting NameNode

STARTUP_MSG:   host = CentOS-Cluster-01/8.8.8.11

STARTUP_MSG:   args = [-format]

STARTUP_MSG:   version = 2.3.0-cdh5.0.1

STARTUP_MSG:   classpath =…

STARTUP_MSG:   build =…

STARTUP_MSG:   java = 1.7.0_60

************************************************************/

 (…略…)

14/06/23 21:02:57 INFO common.Storage: Storage directory /home/zero/hadoop/hadoop-2.3.0-cdh5.0.1/data/dfs/name has been successfully formatted. (…略…)

14/06/23 21:02:59 INFO namenode.NameNode: SHUTDOWN_MSG:

/************************************************************

SHUTDOWN_MSG: Shutting down NameNode at CentOS-Cluster-01/8.8.8.11

************************************************************/

（3）同步NameNode元数据：

同步Hadoop-NN-01元数据到Hadoop-NN-02

主要是：dfs.namenode.name.dir，dfs.namenode.edits.dir还应该确保共享存储目录下(dfs.namenode.shared.edits.dir ) 包含NameNode 所有的元数据。

[[email protected] hadoop-2.0.0-cdh4.4.0]$ scp -r data/ [email protected]:/home/zero/hadoop/hadoop-2.3.0-cdh5.0.1

seen_txid                           100%    2     0.0KB/s   00:00   

VERSION                            100%  201     0.2KB/s   00:00   

seen_txid                           100%    2     0.0KB/s   00:00   

fsimage_0000000000000000000.md5    100%   62     0.1KB/s   00:00   

fsimage_0000000000000000000         100%  121     0.1KB/s   00:00   

VERSION                            100%  201     0.2KB/s   00:00 

（4）初始化ZFCK：

创建ZNode，记录状态信息。

结点Hadoop-NN-01：hdfs zkfc -formatZK

[[email protected] hadoop-2.3.0-cdh5.0.1]$ hdfs zkfc -formatZK

14/06/23 23:22:28 WARN util.NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable

14/06/23 23:22:28 INFO tools.DFSZKFailoverController: Failover controller configured for NameNode NameNode at Hadoop-NN-01/8.8.8.11:8020

14/06/23 23:22:28 INFO zookeeper.ZooKeeper: Client environment:zookeeper.version=3.4.5-cdh5.0.1--1, built on 05/06/2014 18:50 GMT

14/06/23 23:22:28 INFO zookeeper.ZooKeeper: Client environment:host.name=CentOS-Cluster-01

14/06/23 23:22:28 INFO zookeeper.ZooKeeper: Client environment:java.version=1.7.0_60

14/06/23 23:22:28 INFO zookeeper.ZooKeeper: Client environment:java.vendor=Oracle Corporation

14/06/23 23:22:28 INFO zookeeper.ZooKeeper: Client environment:java.home=/usr/runtime/java/jdk1.7.0_60/jre

14/06/23 23:22:28 INFO zookeeper.ZooKeeper: Client environment:java.class.path=...

14/06/23 23:22:29 INFO zookeeper.ZooKeeper: Client environment:java.library.path=/home/zero/hadoop/hadoop-2.3.0-cdh5.0.1/lib/native

14/06/23 23:22:29 INFO zookeeper.ZooKeeper: Client environment:java.io.tmpdir=/tmp

14/06/23 23:22:29 INFO zookeeper.ZooKeeper: Client environment:java.compiler=<NA>

14/06/23 23:22:29 INFO zookeeper.ZooKeeper: Client environment:os.name=Linux

14/06/23 23:22:29 INFO zookeeper.ZooKeeper: Client environment:os.arch=i386

14/06/23 23:22:29 INFO zookeeper.ZooKeeper: Client environment:os.version=2.6.32-431.el6.i686

14/06/23 23:22:29 INFO zookeeper.ZooKeeper: Client environment:user.name=zero

14/06/23 23:22:29 INFO zookeeper.ZooKeeper: Client environment:user.home=/home/zero

14/06/23 23:22:29 INFO zookeeper.ZooKeeper: Client environment:user.dir=/home/zero/hadoop/hadoop-2.3.0-cdh5.0.1

14/06/23 23:22:29 INFO zookeeper.ZooKeeper: Initiating client connection, connectString=Zookeeper-01:2181,Zookeeper-02:2181,Zookeeper-03:2181 sessionTimeout=2000 watcher[email protected]150c2b0

14/06/23 23:22:29 INFO zookeeper.ClientCnxn: Opening socket connection to server CentOS-Cluster-03/8.8.8.13:2181. Will not attempt to authenticate using SASL (unknown error)

14/06/23 23:22:29 INFO zookeeper.ClientCnxn: Socket connection established to CentOS-Cluster-03/8.8.8.13:2181, initiating session

14/06/23 23:22:29 INFO zookeeper.ClientCnxn: Session establishment complete on server CentOS-Cluster-03/8.8.8.13:2181, sessionid = 0x146cc0517b30002, negotiated timeout = 4000

14/06/23 23:22:29 INFO ha.ActiveStandbyElector: Session connected.

14/06/23 23:22:30 INFO ha.ActiveStandbyElector: Successfully created /hadoop-ha/mycluster in ZK.

14/06/23 23:22:30 INFO zookeeper.ZooKeeper: Session: 0x146cc0517b30002 closed

（5）启动

集群启动法：Hadoop-NN-01:

[[email protected] hadoop-2.3.0-cdh5.0.1]$ start-dfs.sh

14/04/23 01:54:58 WARN util.NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable

Starting namenodes on [Hadoop-NN-01 Hadoop-NN-02]

Hadoop-NN-01: starting namenode, logging to /home/puppet/hadoop/cdh4.4/hadoop-2.0.0-cdh4.4.0/logs/hadoop-puppet-namenode-BigData-01.out

Hadoop-NN-02: starting namenode, logging to /home/puppet/hadoop/cdh4.4/hadoop-2.0.0-cdh4.4.0/logs/hadoop-puppet-namenode-BigData-02.out

Hadoop-DN-01: starting datanode, logging to /home/puppet/hadoop/cdh4.4/hadoop-2.0.0-cdh4.4.0/logs/hadoop-puppet-datanode-BigData-03.out

Hadoop-DN-02: starting datanode, logging to /home/puppet/hadoop/cdh4.4/hadoop-2.0.0-cdh4.4.0/logs/hadoop-puppet-datanode-BigData-04.out

Hadoop-DN-03: starting datanode, logging to /home/puppet/hadoop/cdh4.4/hadoop-2.0.0-cdh4.4.0/logs/hadoop-puppet-datanode-BigData-05.out

Starting journal nodes [Hadoop-DN-01 Hadoop-DN-02 Hadoop-DN-03]

Hadoop-DN-01: starting journalnode, logging to /home/puppet/hadoop/cdh4.4/hadoop-2.0.0-cdh4.4.0/logs/hadoop-puppet-journalnode-BigData-03.out

Hadoop-DN-03: starting journalnode, logging to /home/puppet/hadoop/cdh4.4/hadoop-2.0.0-cdh4.4.0/logs/hadoop-puppet-journalnode-BigData-05.out

Hadoop-DN-02: starting journalnode, logging to /home/puppet/hadoop/cdh4.4/hadoop-2.0.0-cdh4.4.0/logs/hadoop-puppet-journalnode-BigData-04.out

14/04/23 01:55:29 WARN util.NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable

Starting ZK Failover Controllers on NN hosts [Hadoop-NN-01 Hadoop-NN-02]

Hadoop-NN-01: starting zkfc, logging to /home/puppet/hadoop/cdh4.4/hadoop-2.0.0-cdh4.4.0/logs/hadoop-puppet-zkfc-BigData-01.out

Hadoop-NN-02: starting zkfc, logging to /home/puppet/hadoop/cdh4.4/hadoop-2.0.0-cdh4.4.0/logs/hadoop-puppet-zkfc-BigData-02.out

单进程启动法：

<1>NameNode（Hadoop-NN-01，Hadoop-NN-02）：

[[email protected] hadoop-2.3.0-cdh5.0.1]$ hadoop-daemon.sh start namenode

<2>DataNode（Hadoop-DN-01，Hadoop-DN-02，Hadoop-DN-03）：

[[email protected] hadoop-2.3.0-cdh5.0.1]$ hadoop-daemon.sh start datanode

<3>JournalNode（Hadoop-DN-01，Hadoop-DN-02，Hadoop-DN-03）：

[[email protected] hadoop-2.3.0-cdh5.0.1]$ hadoop-daemon.sh start journalnode

<4>ZKFC（Hadoop-NN-01，Hadoop-NN-02）：

[[email protected] hadoop-2.3.0-cdh5.0.1]$ hadoop-daemon.sh start zkfc

（6）验证
<1>进程
NameNode：

[[email protected] hadoop-2.3.0-cdh5.0.1]$ jps

4001 NameNode

4290 DFSZKFailoverController

4415 Jps

DataNode：

[[email protected] hadoop-2.3.0-cdh5.0.1]$ jps

25918 QuorumPeerMain

19217 JournalNode

19143 DataNode

19351 Jps

<2>页面:

Active结点:http://8.8.8.10:50070


Hadoop-2.3.0-cdh5.0.1完全分布式环境搭建（NameNode,ResourceManager HA）
 JSP:

Hadoop-2.3.0-cdh5.0.1完全分布式环境搭建（NameNode,ResourceManager HA）
 StandBy结点：

Hadoop-2.3.0-cdh5.0.1完全分布式环境搭建（NameNode,ResourceManager HA）
 JSP:

Hadoop-2.3.0-cdh5.0.1完全分布式环境搭建（NameNode,ResourceManager HA）
 （7）停止:stop-dfs.sh

5）启动Yarn

（1）启动

<1>集群启动

Hadoop-NN-01启动Yarn，命令所在目录：$HADOOP_HOME/sbin

[[email protected] hadoop]$ start-yarn.sh

starting yarn daemons

starting resourcemanager, logging to /home/zero/hadoop/hadoop-2.3.0-cdh5.0.1/logs/yarn-zero-resourcemanager-CentOS-Cluster-01.out

Hadoop-DN-02: starting nodemanager, logging to /home/zero/hadoop/hadoop-2.3.0-cdh5.0.1/logs/yarn-zero-nodemanager-CentOS-Cluster-04.out

Hadoop-DN-03: starting nodemanager, logging to /home/zero/hadoop/hadoop-2.3.0-cdh5.0.1/logs/yarn-zero-nodemanager-CentOS-Cluster-05.out

Hadoop-DN-01: starting nodemanager, logging to /home/zero/hadoop/hadoop-2.3.0-cdh5.0.1/logs/yarn-zero-nodemanager-CentOS-Cluster-03.out

Hadoop-NN-02备机启动RM：

[[email protected] hadoop]$ yarn-daemon.sh start resourcemanager

starting resourcemanager, logging to /home/zero/hadoop/hadoop-2.3.0-cdh5.0.1/logs/yarn-zero-resourcemanager-CentOS-Cluster-02.o

<2>单进程启动

ResourceManager（Hadoop-NN-01，Hadoop-NN-02）：

[[email protected] hadoop-2.3.0-cdh5.0.1]$ yarn-daemon.sh start resourcemanager

DataNode（Hadoop-DN-01，Hadoop-DN-02，Hadoop-DN-03）：

[[email protected] hadoop-2.3.0-cdh5.0.1]$ yarn-daemon.sh start nodemanager

（2）验证

<1>进程：

JobTracker：Hadoop-NN-01，Hadoop-NN-02

[[email protected] hadoop]$ jps

7454 NameNode

14684 Jps

14617 ResourceManager

7729 DFSZKFailoverController

TaskTracker：Hadoop-DN-01，Hadoop-DN-02，Hadoop-DN-03

[[email protected] ~]$ jps

8965 Jps

5228 JournalNode

5137 DataNode

2514 QuorumPeerMain

8935 NodeManager

<2>页面

ResourceManger（Active）：8.8.8.11:23188


Hadoop-2.3.0-cdh5.0.1完全分布式环境搭建（NameNode,ResourceManager HA）
 

ResourceManager（Standby）：8.8.8.12:23188 


Hadoop-2.3.0-cdh5.0.1完全分布式环境搭建（NameNode,ResourceManager HA）
 
Hadoop-2.3.0-cdh5.0.1完全分布式环境搭建（NameNode,ResourceManager HA）
 

（3）停止

Hadoop-NN-01：stop-yarn.sh

Hadoop-NN-02：yarn-daeman.sh stop resourcemanager

来自若泽大数据交流群

相关推荐