一、实验目的:
(2)linux操作系统:centos 6.3 64位
export JAVA_HOME=/usr/local/jdk
---- core-site.xml:
五、格式化namenode和启动hadoop:
(1)格式化namenode
(1)搭建一套hadoop的分布式运行环境,用于学习hadoop相关知识。(2)相对于为分布式和本地模式,分布式环境更接近于实际的生产环境
NameNode JobTracker SecondaryNameNode Slaves chaoren1 chaoren2 chaoren2 chaoren3 、 chaoren4 、 chaoren5
二、实施基础:
(1)阿里云服务器5个节点
节点名称
|
公网
IP
|
内网
IP
|
chaoren1
|
115.29.41.*
|
10.161.74.123
|
chaoren2
|
115.28.240.*
|
10.161.61.164
|
chaoren3
|
115.28.181.*
|
10.144.185.31
|
chaoren4
|
42.96.187.*
|
10.129.89.107
|
chaoren5
|
115.28.196.*
|
10.144.24.126
|
(3)jdk:JAVA_HOME = /usr/local/jdk
(4)以设置免密码登录(参看实验1:ssh免密码登录)
三、实施步骤:
(1)下载hadoop-1.1.2.tar.gz文件并上传至chaoren1 的cr12用户的主目录下的soft文件夹下(如没有需创建:mkdir soft)
(2)解压缩hadoop-1.1.2.tar.gz文件,使用命令:
cd /mnt/home/cr12/softtar -zxvf hadoop-1.1.2.tar.gz
(3)分别修改/mnt/home/cr12/soft/hadoop-1.1.2/conf 下hadoop-env.sh,core-site.xml,hdfs-site.xml,mapred-site.xml 4个文件---- hadoop-env.sh,空白处增加:
export JAVA_HOME=/usr/local/jdk
---- 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://chaoren1:9012</value></property><property><name>hadoop.tmp.dir</name><value>/mnt/home/cr12/soft/hadoop-1.1.2/tmp</value></property>---- hdfs-site.xml:</configuration>
<configuration><property><name>dfs.http.address</name><value>0.0.0.0:50082</value></property><property><name>dfs.secondary.http.address</name><value>0.0.0.0:0</value></property>
<property><name>dfs.datanode.address</name><value>0.0.0.0:0</value></property>
----mapred-site.xml:<property><name>dfs.datanode.ipc.address</name><value>0.0.0.0:0</value></property><property><name>dfs.datanode.http.address</name><value>0.0.0.0:0</value></property><property><name>dfs.replication</name><value>3</value></property><property><name>dfs.permissions</name><value>false</value></property></configuration>
(4)修改master和slaves文件<configuration><property><name>mapred.job.tracker</name><value>http://chaoren2:8012</value></property><property><name>mapred.job.tracker.http.address</name><value>0.0.0.0:50042</value></property><property><name>mapred.task.tracker.http.address</name><value>0.0.0.0:0</value></property></configuration>
----master文件增加如下内容,指定jobtracker主机为chaoren2:
chaoren2
四、将hadoop文件分发到个节点中去:
----slaves文件增加如下内容,指定从节点:
chaoren3
chaoren4
chaoren5
(1)在chaoren1上执行:
scp -r ~/soft chaoren2:~scp -r ~/soft chaoren3:~scp -r ~/soft chaoren4:~scp -r ~/soft chaoren5:~
五、格式化namenode和启动hadoop:
(1)格式化namenode
1、cd到hadoop文件夹下,执行bin/hadoop namenode -format ,如提示format successfuly,表示格式化成功2、如格式化过程中报错,仔细看报错信息或查看hadoopm目录下的log文件信息,定位错误并修改,再次执行格式化命令
(2)启动hadoop
----chaoren1下启动namenode(3)如何验证是否启动成功?
bin/hadoop-daemon.sh start namenode
----chaoren2下启动SecondaryNamenode和JobTracker
bin/hadoop-daemon.sh start secondarynamenode
bin/hadoop-daemon.sh start jobtracker
----chaoren1下启动Datanode和TaskTracker
bin/hadoop-daemons.sh satrt datanode
bin/hadoop-daemons.sh start tasktracker
1、在各节点上输入jps命令查询是否有相应的进程2、尝试访问浏览器中访问chaoren1:50082 和 chaoren2:50042
(4)关闭hadoop集群
----chaoren1下关闭Datanode和TaskTracker
bin/hadoop-daemons.sh stop datanode
bin/hadoop-daemons.sh stop tasktracker
----chaoren2下关闭SecondaryNamenode和JobTracker
bin/hadoop-daemon.sh stop secondarynamenode
bin/hadoop-daemon.sh stop jobtracker
----chaoren1下关闭namenode
bin/hadoop-daemon.sh stop namenode