[转]hadoop-2.5安装与配置

安装以前准备4台机器:bluejoe0,bluejoe4,bluejoe5,bluejoe9html

bluejoe0做为master,bluejoe4,5,9做为slavejava

bluejoe0做为namenodenode

bluejoe9为secondary namenodeweb

bluejoe4,5,9做为datanodeapache

安装hadoop

首先在bluejoe0机器上下载Hadoop浏览器

wget http://mirrors.cnnic.cn/apache/hadoop/common/stable2/hadoop-2.5.2.tar.gzapp

保存至/usr/local/,tar之;框架

ln之,/usr/local/hadoop;webapp

配置hdfs

配置core-site.xml:jvm

 

[html] view plain copy

 print?在CODE上查看代码片派生到个人代码片

  1. <configuration>  
  2.         <property>     
  3.                 <name>fs.default.name</name>     
  4.                 <value>hdfs://bluejoe0:9000</value>     
  5.         </property>  
  6.         <property>  
  7.                 <name>io.file.buffer.size</name>  
  8.                 <value>4096</value>  
  9.         </property>  
  10.   
  11.         <property>  
  12.                  <name>hadoop.tmp.dir</name>  
  13.                  <value>/data/hdfs/tmp</value>  
  14.         </property>  
  15. </configuration>  


配置hdfs-site.xml:

 

 

[html] view plain copy

 print?在CODE上查看代码片派生到个人代码片

  1. <?xml version="1.0" encoding="UTF-8"?>  
  2. <?xml-stylesheet type="text/xsl" href="configuration.xsl"?>  
  3.   
  4. <configuration>  
  5.         <property>    
  6.                 <name>dfs.name.dir</name>    
  7.                 <value>file:/data/hdfs/name</value>    
  8.         </property>    
  9.         <property>    
  10.                 <name>dfs.data.dir</name>     
  11.                 <value>file:/data/hdfs/data</value>    
  12.         </property>    
  13.         <property>    
  14.                 <name>dfs.permissions</name>    
  15.                 <value>false</value>    
  16.        </property>  
  17.         <property>  
  18.                 <name>dfs.replication</name>  
  19.                 <value>2</value>  
  20.         </property>  
  21.         <property>  
  22.                 <name>dfs.webhdfs.enabled</name>  
  23.                 <value>true</value>  
  24.         </property>  
  25.         <property>  
  26.                 <name>dfs.namenode.rpc-address</name>  
  27.                 <value>bluejoe0:9000</value>  
  28.         </property>  
  29.         <property>  
  30.                 <name>dfs.namenode.secondary.http-address</name>  
  31.                 <value>bluejoe9:50090</value>  
  32.         </property>  
  33. </configuration>  

 

注意,dfs.namenode.rpc-address要和fs.default.name一致。

 

设置/usr/local/hadoop/etc/hadoop/hadoop-env.sh:

export JAVA_HOME=/usr/lib/jvm/Java-1.6.0-openjdk-1.6.0.0.x86_64

 

scp,将hadoop目录复制到其它机器;

 

配置slaves:

 

[html] view plain copy

 print?在CODE上查看代码片派生到个人代码片

  1. bluejoe4  
  2. bluejoe5  
  3. bluejoe9  

 

namenode格式化:

hdfs namenode -format

启动hdfs:

./sbin/start-dfs.sh

能够看到输出信息:

 

[html] view plain copy

 print?在CODE上查看代码片派生到个人代码片

  1. Starting namenodes on [bluejoe0]  
  2. bluejoe0: starting namenode, logging to /usr/local/hadoop-2.5.2/logs/hadoop-root-namenode-bluejoe0.out  
  3. bluejoe9: starting datanode, logging to /usr/local/hadoop-2.5.2/logs/hadoop-root-datanode-bluejoe9.out  
  4. bluejoe4: starting datanode, logging to /usr/local/hadoop-2.5.2/logs/hadoop-root-datanode-bluejoe4.out  
  5. bluejoe5: starting datanode, logging to /usr/local/hadoop-2.5.2/logs/hadoop-root-datanode-bluejoe5.out  
  6. Starting secondary namenodes [bluejoe9]  


 

 

接下来,能够查看Web界面(http://bluejoe0:50070/),其datanodes截图以下:

目前为止,hdfs安装完毕!

配置mapreduce

修改 yarn-site.xml:

 

[html] view plain copy

 print?在CODE上查看代码片派生到个人代码片

  1. <?xml version="1.0"?>  
  2. <configuration>  
  3.         <property>  
  4.                 <name>yarn.nodemanager.aux-services</name>  
  5.                 <value>mapreduce_shuffle</value>  
  6.         </property>  
  7.   
  8.         <property>  
  9.                 <name>yarn.nodemanager.aux-services.mapreduce.shuffle.class</name>  
  10.                 <value>org.apache.hadoop.mapred.ShuffleHandler</value>  
  11.         </property>  
  12.   
  13.         <property>  
  14.                 <name>yarn.resourcemanager.address</name>  
  15.                 <value>bluejoe0:8032</value>  
  16.         </property>  
  17.   
  18.         <property>  
  19.                 <name>yarn.resourcemanager.scheduler.address</name>  
  20.                 <value>bluejoe0:8030</value>  
  21.         </property>  
  22.   
  23.         <property>  
  24.                 <name>yarn.resourcemanager.resource-tracker.address</name>  
  25.                 <value>bluejoe0:8031</value>  
  26.          </property>  
  27.   
  28.          <property>  
  29.                 <name>yarn.resourcemanager.admin.address</name>  
  30.                 <value>bluejoe0:8033</value>  
  31.         </property>  
  32.   
  33.         <property>  
  34.                 <name>yarn.resourcemanager.webapp.address</name>  
  35.                 <value>bluejoe0:8088</value>  
  36.         </property>  
  37. </configuration>  


scp如上配置文件至其它节点;

 

启动mapreduce框架:

/usr/local/hadoop-2.5.2/sbin/start-yarn.sh

启动浏览器,访问http://bluejoe0:8088:

执行测试程序:

hadoop jar /usr/local/hadoop/share/hadoop/mapreduce/hadoop-mapreduce-examples-2.5.2.jar pi 100 1000

Job Finished in 12.885 seconds
Estimated value of Pi is 3.14120000000000000000

设置/usr/local/hadoop/etc/hadoop/hadoop-env.sh:

export JAVA_HOME=/usr/lib/jvm/java-1.6.0-openjdk-1.6.0.0.x86_64

相关文章
相关标签/搜索