hadoop安装（hadoop2.0.4alpha）

安装环境&＃xff1a;centos ;

             jdk-7-linux-x64.tar.gz

             hadoop-2.0.4-alpha.tar.gz

安装目录&＃xff1a;/opt/cloud

1、首先安装jdk&＃xff1a;

    tar -zvxf  jdk-7-linux-x64.tar.gz&＃xff0c;将jdk解压至 /opt/cloud/jdk&＃xff0c;设置环境变量&＃xff0c;亦可不设置。

2、解压 hadoop-2.0.4-alpha.tar.gz

    tar -zvxf hadoop-2.0.4-alpha.tar.gz&＃xff0c;将hadoop解压至 /opt/cloud/hadoop&＃xff0c;可修改目录或者软连接。

3、配置 hadoop

• ssh 免密码登陆&＃xff1a;ssh-keygen -t rsa&＃xff0c;使用 ssh localhost 测试&＃xff0c;直接进入ssh则成功
• Hadoop 环境变量配置

• 修改Hadoop的配置文件&＃xff1a;
  hadoop-env.sh&＃xff1a;
  #vim /opt/cloud/hadoop/etc/hadoop/hadoop-env.sh

    修改 export JAVA_HOME&＃61;/opt/cloud/jdk

• 
  

  编辑以下几个文件&＃xff0c;加入配置信息&＃xff0c;文件位于 hadoop/etc/hadoop

  ----------------core-site.xml

  
  
  
  
  fs.default.name
  hdfs://localhost:8020 
  The name of the default file system. Either the literal string "local" or a host:port for NDFS.
  true 

  
  

  
  
  
  
  ------------------------- yarn-site.xml
  
  
  
  
   

  yarn.nodemanager.aux-services 

  mapreduce.shuffle 

   

  
  

  
  

   

  yarn.nodemanager.aux-services.mapreduce.shuffle.class 

  org.apache.hadoop.mapred.ShuffleHandler 

   
  ------------------------ mapred-site.xml
  
  
   

  mapreduce.framework.name 

  yarn 

   

  
  

  
  

   

  mapred.system.dir 

  file:/opt/cloud/hadoop_space/mapred/system 

  true 

   

  
  

  
  

   

  mapred.local.dir 

  file:/opt/cloud/hadoop_space/mapred/local 

  true 

   
  
  
  
  
  
  
  ----------- hdfs-site.xml  
  
  
  
  
  dfs.namenode.name.dir

  file:/opt/cloud/hadoop_space/dfs/name

  Determines where on the local filesystem the DFS name node should store 

  the name table. If this is a comma-delimited list 

  of directories then the name table is replicated in all of the directories, for redundancy.

  true

  
  

  dfs.datanode.data.dir

  file:/opt/cloud/hadoop_space/dfs/data

  Determines where on the local 

  filesystem an DFS data node should store its blocks. If this is a comma-delimited 

  list of directories, then data will be stored in all named 

  directories, typically on different devices. 

  Directories that do not exist are ignored. 

  true

  dfs.replication

  1

  
  

  
  

  dfs.permissions

  false

4、测试

以上配置好后启动HDFS&＃xff1a;
# hdfs namenode -format 
运行成功以后可以使用一下命令启动NameNode和DataNode
# hadoop-daemon.sh start namenode 

# hadoop-daemon.sh start datanode 

打开&＃xff1a;http://mycentos:50070/dfshealth.jsp