单节点Hadoop安装过程

1.1.1 环境准备

本次由一台centos虚拟服务器搭建hadoop平台，机器信息如表１所示：

表１ 主机环境准备

名称	信息
IP	10.1.1.20
hostname	Master.hadoop

为方便使用，现给出主机环境方面需要修改的地方：

●　IP地址修改

IP地址位于/etc/sysconfig/network-scripts/目录中，通过vi编辑ifcfg-eth0文件修改成如下所示结构即可：

[root@master network-scripts]# cd /etc/sysconfig/network-scripts/

[root@master network-scripts]# cat ifcfg-eth0

DEVICE="eth0"

ONBOOT=yes

TYPE=Ethernet

BOOTPROTO=none

IPADDR=10.1.1.20

PREFIX=24

GATEWAY=10.1.1.1

DEFROUTE=yes

HWADDR=00:30:16:AF:00:D1

●　hostname修改

Hostname修改位于/etc/sysconfig/network文件中，其修改后的结果如下所示：

[root@master network-scripts]# cat /etc/sysconfig/network

NETWORKING=yes

HOSTNAME=master.hadoop

●　DNS修改

DNS位于/etc/hosts文件中，修改的结果如下所示：

[root@master network-scripts]# cat /etc/hosts

10.1.1.20master.hadoop master

127.0.0.1 localhost.localdomain localhost

●　环境测试

通过ping测试master.hadoop是否畅通：

[root@master network-scripts]# ping master.hadoop

PING master.hadoop (10.1.1.20) 56(84) bytes of data.

64 bytes from master.hadoop (10.1.1.20): icmp_seq=1 ttl=64 time=0.040 ms

64 bytes from master.hadoop (10.1.1.20): icmp_seq=2 ttl=64 time=0.016 ms

--- master.hadoop ping statistics ---

2 packets transmitted, 2 received, 0% packet loss, time 1467ms

rtt min/avg/max/mdev = 0.016/0.028/0.040/0.012 ms

1.1.2Java安装与部署

Hadoop需要java环境支持，通常需要java 1.6版本以上，因此可以通过去java官方网站下载JDK环境，下载地址为：

http://www.oracle.com/technetwork/java/javase/downloads/jdk-6u25-download-346242.html

从本链接中选择jdk-6u25-linux-x64-rpm.bin，在接受协议后方可下载到本地；

　 ●　Java安装

将下载到后java文件传至master.hadoop主机/home目录中，下面可以进行对其进行安装：

[root@master home]# chmod u+x jdk-6u25-linux-x64-rpm.bin

[root@master home]# ./jdk-6u25-linux-x64-rpm.bin

●　Java配置

　　Java安装完毕后，可以对java目录信息进行环境变量配置，配置信息需增加至文件/etc/profile之中，具体如下所示：

[root@master home]#vi /etc/profile

JAVA_HOME=/usr/java/jdk1.6.0_25

CLASSPATH=.:$JAVA_HOME/lib

PATH=$PATH:$JAVA_HOME/bin:$JAVA_HOME/jre/bin

　

环境变量配置完毕后，通过命令进行检验并生效：　

[root@master jdk1.6.0_25]# source /etc/profile

1.1.3 SSH配置

通过配置SSH实现基于公钥方式无密码登录，具体操作步骤为：创建一个新的hadoop帐户、生成这个帐户的SSH公钥、配置公钥授权文件、设置SSH服务登录方式等，下面给出具体方式：

●　创建hadoop帐户

[root@master jdk1.6.0_25]# useradd hadoop #创建帐号

[root@master jdk1.6.0_25]# passwd hadoop #配置密码

●　生成公钥

[hadoop@master ~]$ ssh-keygen 　　　#生成SSH认证公钥，连续回车即可

Generating public/private rsa key pair.

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

Enter passphrase (empty for no passphrase):

Enter same passphrase again:

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

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

The key fingerprint is:

86:b5:d9:6a:ea:03:4e:5a:97:e5:24:5b:1f:65:41:89 [e-mail]

The key's randomart image is:

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

| ooo |

| E + |

| . o |

| .o++. |

| .OS... |

| + +.... |

| = o o |

| . . .o |

| .o. |

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

[hadoop@master ~]$ cd .ssh/

[hadoop@master .ssh]$ ls

id_rsa id_rsa.pub

●　配置授权

[hadoop@master ~]$ cat ~/.ssh/id_rsa.pub >> ~/.ssh/authorized_keys

[hadoop@master ~]$ chmod 700 ~/.ssh

[hadoop@master ~]$ chmod 600 ~/.ssh/authorized_keys

● 测试

[hadoop@master jdk1.6.0_25]$ ssh master.hadoop

Last login: Wed Jun 13 18:29:29 2012 from master.hadoop

1.1.4 Hadoop安装与配置

使用的Hadoop版本是hadoop-0.20.2.tar.gz，下载地址为：

http://mirror.bit.edu.cn/apache/hadoop/common/hadoop-0.20.2/hadoop-0.20.2.tar.gz

　 ●　Hadoop安装

[root@master home]# tar xzvf hadoop-0.20.2.tar.gz

[root@master home]# mv hadoop-0.20.2 /usr/local

[root@master home]# cd /usr/local

[root@master local]# ls

bin etc games hadoop-0.20.2 include lib lib64 libexec sbin share src

[root@master local]# mv hadoop-0.20.2/ hadoop

[root@master local]# ls

bin etc games hadoop include lib lib64 libexec sbin share src

[root@master local]# chown -R hadoop:hadoop /usr/local/hadoop/　#修改权限

●　环境变量配置

跟配置JAVA一样配置hadoop环境变量，编辑文件/etc/profile，同时也要修改hadoop内部环境变量/hadoop/conf/hadoop_env.sh，具体细节如下所示：

[root@master local]# vi /etc/profile

HADOOP_HOME=/usr/local/hadoop

HADOOP_CONF_DIR=$HADOOP_HOME/conf

CLASSPAH=.:$JAVA_HOME/lib:$HADOOP_HOME/lib

PATH=$PATH:$JAVA_HOME/bin:$JAVA_HOME/jre/bin:$HADOOP_HOME/bin

"/etc/profile" 73L, 1660C written

[root@master local]# source /etc/profile

[root@master conf]# vi hadoop-env.sh

export JAVA_HOME=$JAVA_HOME
export HADOOP_CLASSPATH="$HADOOP_CLASSPATH"
export HADOOP_HEAPSIZE=2048
export HADOOP_LOG_DIR=/var/local/logs
export HADOOP_PID_DIR=/var/local/pids

[root@master bin]# export JAVA_HOME

[root@master bin]# export HADOOP_HOME

[root@master bin]# export HADOOP_CONF_DIR

●　hadoop文件配置

配置三个xml文件，分别为：core-site.xml、hdfs-site.xml、mapred-site.xml，配置效果如下所示：

文件：core-site.xml

[root@master conf]# vi core-site.xml

<?xml version="1.0"?>

<?xml-stylesheet type="text/xsl" href="configuration.xsl"?>

<!-- Put site-specific property overrides in this file. -->

<configuration>

<property>

<name>fs.default.name</name>

<value>hdfs://localhost:9000</value>

</property>

</configuration>

文件：hdfs-site.xml

[root@master conf]# vi hdfs-site.xml

<?xml version="1.0"?>

<?xml-stylesheet type="text/xsl" href="configuration.xsl"?>

<!-- Put site-specific property overrides in this file. -->

<configuration>

<property>

<name>dfs.replication</name>

<value>1</value>

</property>

</configuration>

文件：mapred-site.xml

[root@master conf]# vi mapred-site.xml

<?xml version="1.0"?>

<?xml-stylesheet type="text/xsl" href="configuration.xsl"?>

<!-- Put site-specific property overrides in this file. -->

<configuration>

<property>

<name>mapred.job.tracker</name>

<value>localhost:9001</value>

</property>

</configuration>

●　hadoop格式化文件系统

　切换到bin目录，找到可执行文件hadoop，执行文件系统格式化操作：

[root@master bin]# hadoop namenode -format

●　启动hadoop

[root@master bin]# ./start-all.sh

starting namenode, logging to /var/local/logs/hadoop-root-namenode-master.hadoop.out

localhost: starting datanode, logging to /var/local/logs/hadoop-root-datanode-master.hadoop.out

localhost: starting secondarynamenode, logging to /var/local/logs/hadoop-root-secondarynamenode-master.hadoop.out

starting jobtracker, logging to /var/local/logs/hadoop-root-jobtracker-master.hadoop.out

localhost: starting tasktracker, logging to /var/local/logs/hadoop-root-tasktracker-master.hadoop.out

1.1.5Hadoop测试

[root@master hadoop]# jps

2459 JobTracker

2284 DataNode

2204 NameNode

2860 Jps

2382 SecondaryNameNode

2575 TaskTracker