CDH5 完美手动配置过程改进版

一、安装前准备：
操作系统：CentOS 6.5 64位操作系统
环境：jdk1.7.0_45以上，本次采用jdk-7u55-linux-x64.tar.gz
master01 10.10.2.57 namenode 节点
master02 10.10.2.58 namenode 节点
slave01:10.10.2.173 datanode 节点
slave02:10.10.2.59 datanode 节点
slave03: 10.10.2.60 datanode 节点
注：Hadoop2.0以上采用的是jdk环境是1.7,Linux自带的jdk卸载掉，重新安装
下载地址：http://www.oracle.com/technetwork/java/javase/downloads/index.html
软件版本：hadoop-2.3.0-cdh5.1.0.tar.gz, zookeeper-3.4.5-cdh5.1.0.tar.gz
下载地址：http://archive.cloudera.com/cdh5/cdh/5/
开始安装：
二、jdk安装
1、检查是否自带jdk
rpm -qa | grep jdk
java-1.6.0-openjdk-1.6.0.0-1.45.1.11.1.el6.i686 
2、卸载自带jdk
yum -y remove java-1.6.0-openjdk-1.6.0.0-1.45.1.11.1.el6.i686
3、安装jdk-7u55-linux-x64.tar.gz
在usr/目录下创建文件夹java,在java文件夹下运行tar –zxvf jdk-7u55-linux-x64.tar.gz
解压到java目录下
[root@master01 java]# ls
jdk1.7.0_55
三、配置环境变量
远行vi /etc/profile
# /etc/profile
# System wide environment and startup programs, for login setup
# Functions and aliases go in /etc/bashrc
export JAVA_HOME=/usr/java/jdk1.7.0_55
export JRE_HOME=/usr/java/jdk1.7.0_55/jre
export CLASSPATH=/usr/java/jdk1.7.0_55/lib
export PATH=$JAVA_HOME/bin: $PATH
保存修改，运行source /etc/profile 重新加载环境变量
运行java -version
[root@master01 java]# java -version
java version "1.7.0_55"
Java(TM) SE Runtime Environment (build 1.7.0_55-b13)
Java HotSpot(TM) 64-Bit Server VM (build 24.55-b03, mixed mode)
Jdk配置成功
四、系统配置
预先准备5台机器，并配置IP
关闭防火墙
chkconfig iptables off(永久性关闭)
配置主机名和hosts文件
[root@master01 java]# vi /etc/hosts
127.0.0.1   localhost localhost.localdomain localhost4 localhost4.localdomain4
::1         localhost localhost.localdomain localhost6 localhost6.localdomain6
10.10.2.57 master01
10.10.2.58 master02
10.10.2.173 slave01
10.10.2.59 slave02
10.10.2.60 slave03
按照不同机器IP配置不同的主机名
3、SSH无密码验证配置
因为Hadoop运行过程需要远程管理Hadoop的守护进程，NameNode节点需要通过SSH（Secure Shell）链接各个DataNode节点，停止或启动他们的进程，所以SSH必须是没有密码的，所以我们要把NameNode节点和DataNode节点配制成无秘密通信，同理DataNode也需要配置无密码链接NameNode节点。
在每一台机器上配置：
vi /etc/ssh/sshd_config打开
RSAAuthentication yes # 启用 RSA 认证，PubkeyAuthentication yes # 启用公钥私钥配对认证方式
Master01:运行：ssh-keygen –t rsa –P ''  不输入密码直接enter
默认存放在 ／root／.ssh目录下,
cat ~/.ssh/id_rsa.pub >> ~/.ssh/authorized_keys
[root@master01 .ssh]# ls
authorized_keys  id_rsa  id_rsa.pub  known_hosts
slave01执行相同的操作，然后将master01 ／root／.ssh/目录下的id_rsa.pub放到 slave01 相同目录下的authorized_keys这样slave01就持有了master01的公钥 然后直接ssh slave01测试是否可以无密码连接到slave01上，然后将slave01 上的id_rsa.pub 追加到master01的authorized_keys中，测试ssh master01 是否可以直接连上slave01.
[root@master01 ~]# ssh slave01
Last login: Tue Aug 19 14:28:15 2014 from master01
[root@slave01 ~]# 
Master01-master02
Master01-slave01
Master01-slave02
Master01-slave03
Master02-slave01
Master02-slave02
Master02-slave03
执行相同的操作。
  
五、安装Hadoop
建立文件目录 /usr/local/cloud 创建文件夹data，存放数据、日志文件，haooop原文件，zookeeper原文件
[root@slave01 cloud]# ls
data  hadoop  tar  zookeeper
5.1、配置hadoop-env.sh
进入到/usr/local/cloud/hadoop/etc/hadoop目录下
配置vi hadoop-env.sh hadoop运行环境加载
export JAVA_HOME=/usr/java/jdk1.7.0_55
5.2、配置core-site.xml


    hadoop.tmp.dir
    /usr/local/cloud/data/hadoop/tmp



    fs.defaultFS
    hdfs://zzg


 
    ha.zookeeper.quorum
    master01:2181,slave01:2181,slave02:2181

  
(2)hdfs-site.xml配置


    dfs.namenode.name.dir
    /usr/local/cloud/data/hadoop/dfs/nn



    dfs.datanode.data.dir
    /usr/local/cloud/data/hadoop/dfs/dn



    dfs.replication
    3



    dfs.webhdfs.enabled
    true



     dfs.permissions
     false



     dfs.permissions.enabled
     false




    dfs.nameservices
    zzg



    dfs.ha.namenodes.zzg
    nn1,nn2



    dfs.namenode.rpc-address.zzg.nn1
    master01:9000


    dfs.namenode.rpc-address.zzg.nn2
    master02:9000



    dfs.namenode.http-address.zzg.nn1
    master01:50070


    dfs.namenode.http-address.zzg.nn2
    master02:50070



    dfs.namenode.servicerpc-address.zzg.nn1
    master01:53310


    dfs.namenode.servicerpc-address.zzg.nn2
    master02:53310



    dfs.namenode.shared.edits.dir
    qjournal://master01:8485;slave01:8485;slave02:8485/zzg

 

    dfs.journalnode.edits.dir
    /usr/local/cloud/data/hadoop/ha/journal



    dfs.client.failover.proxy.provider.zzg
    org.apache.hadoop.hdfs.server.namenode.ha.ConfiguredFailoverProxyProvider



    dfs.ha.automatic-failover.enabled
    true


        ha.zookeeper.quorum
        master01:2181,slave01:2181,slave02:2181



    dfs.ha.fencing.methods
    sshfence



    dfs.ha.fencing.ssh.private-key-files
    /root/.ssh/id_rsa

5.3 配置maped-site.xml

                mapreduce.framework.name
                yarn

5.4配置yarn HA 
配置yarn-en.sh java环境
# some Java parameters
  export JAVA_HOME=/usr/java/jdk1.7.0_55
5.5配置yarn-site.xml
        
        
                yarn.resourcemanager.connect.retry-interval.ms
                2000
        
        
         
                yarn.resourcemanager.ha.enabled
                true
        
        
        
                yarn.resourcemanager.ha.automatic-failover.enabled
                true
        
        
        
                yarn.resourcemanager.ha.rm-ids
                rm1,rm2
        
        
        
                yarn.resourcemanager.ha.id
                rm1
               If we want to launch more than one RM in single node, we need this configuration
         
        
         
                yarn.resourcemanager.recovery.enabled
                 true
        
        
        
                yarn.resourcemanager.zk-state-store.address
                localhost:2181
        
  
        
                yarn.resourcemanager.store.class
                org.apache.hadoop.yarn.server.resourcemanager.recovery.ZKRMStateStore
        
        
                yarn.resourcemanager.zk-address
                localhost:2181
        
        
                yarn.resourcemanager.cluster-id
                yarn-cluster
        
        
         
                yarn.app.mapreduce.am.scheduler.connection.wait.interval-ms
                5000
        
        
        
                yarn.resourcemanager.address.rm1
                master01:23140
        
        
                yarn.resourcemanager.scheduler.address.rm1
                master01:23130
        
        
                yarn.resourcemanager.webapp.address.rm1
                master01:23188
        
        
                yarn.resourcemanager.resource-tracker.address.rm1
                master01:23125
        
         
                yarn.resourcemanager.admin.address.rm1
                master01:23141
        
        
                yarn.resourcemanager.ha.admin.address.rm1
                master01:23142
        
        
         
                yarn.resourcemanager.address.rm2
                master02:23140
        
        
                yarn.resourcemanager.scheduler.address.rm2
                master02:23130
        
        
                yarn.resourcemanager.webapp.address.rm2
                master02:23188
        
        
                yarn.resourcemanager.resource-tracker.address.rm2
                master02:23125
        
        
                yarn.resourcemanager.admin.address.rm2
                master02:23141
        
        
                yarn.resourcemanager.ha.admin.address.rm2
                master02:23142
        
        
        
                Address where the localizer IPC is.
                yarn.nodemanager.localizer.address
                0.0.0.0:23344
        
        
         
                NM Webapp address.
                yarn.nodemanager.webapp.address
                0.0.0.0:23999
        
        
                yarn.nodemanager.aux-services
                mapreduce_shuffle
        
        
                yarn.nodemanager.aux-services.mapreduce.shuffle.class
                org.apache.hadoop.mapred.ShuffleHandler
        
        
                yarn.nodemanager.local-dirs
                /usr/local/cloud/data/hadoop/yarn/local
        
        
                yarn.nodemanager.log-dirs
                /usr/local/cloud/data/logs/hadoop
        
        
                mapreduce.shuffle.port
                23080
        
        
         
                yarn.client.failover-proxy-provider
                 org.apache.hadoop.yarn.client.ConfiguredRMFailoverProxyProvider
         
六、配置zookeeper集群
在zookeeper目录下建立data目录 和logs目录，
配置zoo.cnf
dataDir=/usr/local/cloud/zookeeper/data
dataLogDir=/usr/local/cloud/zookeeper/logs
# the port at which the clients will connect
clientPort=2181
server.1=master01:2888:3888
server.2=master02:2888:3888
server.3=slave01:2888:3888
server.4=slave02:2888:3888
server.5=slave03:2888:3888
在data目录下创建myid文件，并在对应的机器上填写数字，如上配置master01 server01 的myid写入1，
master02 中的data的myid写入2，依次在其他机子上执行相同操作。
在各个机器下zookeeper目录下的bin目录下执行zkServer.sh start命令
再运行zkServer.sh status如果出现leader 或fllower 则说明集群配置正确。
  
到此各个配置文件配置完毕
七、启动Hadoop集群严格按照以下顺序执行（第一次）
(1)各个节点启动zookeeper，在zookeeper/bin/zkServer.sh start
(2)任选一个NN执行完成即可 在hadoop/bin/ hdfs zkfc –formatZK 进行格式化创建命名空间
(3)在配置了journalnode的节点启动，master01，slave01,slave02
   在hadoop/sbin/hadoop-daemon.sh start journalnode
（4）在主namenode节点执行格式化
 hadoop namenode -format zzg
 主机器上启动namenode
 hadoop/sbin/ hadoop-daemon.sh start namenode
(5)将主namenode节点格式化的目录拷贝到从主namenode节点上
hadoop/bin/hdfs namenode –bootstrapStandby
hadoop/sbin/hadoop-daemon.sh start namenode
(6) 在两个namenode节点都执行以下命令
./sbin/hadoop-daemon.sh start zkfc
(7) 在所有datanode节点都执行以下命令启动datanode
$HADOOP_HOME/sbin/hadoop-daemon.sh start datanode
(8)在主namenode节点启动yarn，运行yarn-start.sh命令
jps可以看到
namenode节点
[root@master01 ~]# jps
38972 JournalNode
38758 NameNode
39166 DFSZKFailoverController
37473 QuorumPeerMain
39778 ResourceManager
42620 Jps
datanode节点
[root@slave01 ~]# jps
33440 DataNode
35277 Jps
32681 QuorumPeerMain
33568 JournalNode
34231 NodeManager