hadoop 1.x 伪分布式搭建

下载hadoop1.0.4版本，和jdk1.6版本或更高版本：
1. 安装JDK，安装目录大家可以自定义，下面是我的安装目录：

1. /usr/jdk1.6.0_22

复制代码

配置环境变量：

1. [root@hadoop hadoop-1.0.4]# vi /etc/profile

复制代码

将环境变量添加到profile文件底部：

1. export JAVA_HOME=/usr/jdk1.6.0_22
   
2. export PATH=$PATH:$JAVA_HOME/bin

复制代码

激活环境变量：

1. [root@hadoop hadoop-1.0.4]# source /etc/profile

复制代码

2. 安装hadoop
创建hadoop用户及其hadoop组。
配置IP对应域名在

1. [root@hadoop hadoop-1.0.4]$ vi /etc/hosts

复制代码

在hosts文件底部中添加如下：

1. 192.168.0.101   hadoop.master

复制代码

接下来下载hadoop安装包，大家可以到apache官网下载1.0.4版本
我这里将hadoop安装到hadoop用户目录下，解压hadoop安装文件，hadoop有三个核心配置文件core-site.xml,hdfs-site.xml,mapred-site.xml，
配置core-site.xml

1. <?xml version="1.0"?>
   
2. <?xml-stylesheet type="text/xsl" href="configuration.xsl"?>
   
3. 
   
4. <!-- Put site-specific property overrides in this file. -->
   
5. 
   
6. <configuration>
   
7. <property>
   
8.   <name>fs.default.name</name>
   
9.   <value>hdfs://hadoop.master:9000</value>
   
10. </property>
    
11. <property>
    
12.   <name>hadoop.tmp.dir</name>
    
13.   <value>/home/hadoop/tmp</value>
    
14. </property>
    
15. </configuration>

复制代码

配置hdfs-site.xml

1. <?xml version="1.0"?>
   
2. <?xml-stylesheet type="text/xsl" href="configuration.xsl"?>
   
3. 
   
4. <!-- Put site-specific property overrides in this file. -->
   
5. 
   
6. <configuration>
   
7. <property>
   
8.    <name>dfs.replication</name>
   
9.    <value>1</value>
   
10. </property>
    
11. <property>
    
12.   <name>dfs.name.dir</name>
    
13.   <value>/home/hadoop/hdfs/name</value>
    
14. </property>
    
15. <property>
    
16.   <name>dfs.data.dir</name>
    
17.   <value>/home/hadoop/hdfs/data</value>
    
18. </property>
    
19. </configuration>

复制代码

配置 mapred-site.xml

1. <?xml version="1.0"?>
   
2. <?xml-stylesheet type="text/xsl" href="configuration.xsl"?>
   
3. 
   
4. <!-- Put site-specific property overrides in this file. -->
   
5. 
   
6. <configuration>
   
7. <property>
   
8.    <name>mapred.job.tracker</name>
   
9.    <value>hadoop.master:9001</value>
   
10. </property>
    
11. </configuration>

复制代码

最后配置haoop-env.sh文件

1. # Set Hadoop-specific environment variables here.
   
2. 
   
3. # The only required environment variable is JAVA_HOME.  All others are
   
4. # optional.  When running a distributed configuration it is best to
   
5. # set JAVA_HOME in this file, so that it is correctly defined on
   
6. # remote nodes.
   
7. 
   
8. # The java implementation to use.  Required.
   
9. export JAVA_HOME=/usr/jdk1.6.0_22
   
10. 
    
11. # Extra Java CLASSPATH elements.  Optional.
    
12.   export HADOOP_HOME=/home/hadoop/hadoop-1.0.4
    
13. # export HADOOP_CLASSPATH=
    
14. 
    
15. # The maximum amount of heap to use, in MB. Default is 1000.
    
16. # export HADOOP_HEAPSIZE=2000
    
17. 
    
18. # Extra Java runtime options.  Empty by default.
    
19. # export HADOOP_OPTS=-server
    
20. 
    
21. # Command specific options appended to HADOOP_OPTS when specified
    
22. export HADOOP_NAMENODE_OPTS="-Dcom.sun.management.jmxremote $HADOOP_NAMENODE_OPTS"
    
23. export HADOOP_SECONDARYNAMENODE_OPTS="-Dcom.sun.management.jmxremote $HADOOP_SECONDARYNAMENODE_OPTS"
    
24. export HADOOP_DATANODE_OPTS="-Dcom.sun.management.jmxremote $HADOOP_DATANODE_OPTS"
    
25. export HADOOP_BALANCER_OPTS="-Dcom.sun.management.jmxremote $HADOOP_BALANCER_OPTS"
    
26. export HADOOP_JOBTRACKER_OPTS="-Dcom.sun.management.jmxremote $HADOOP_JOBTRACKER_OPTS"
    
27. # export HADOOP_TASKTRACKER_OPTS=
    
28. # The following applies to multiple commands (fs, dfs, fsck, distcp etc)
    
29. # export HADOOP_CLIENT_OPTS
    
30. 
    
31. # Extra ssh options.  Empty by default.
    
32. # export HADOOP_SSH_OPTS="-o ConnectTimeout=1 -o SendEnv=HADOOP_CONF_DIR"
    
33. 
    
34. # Where log files are stored.  $HADOOP_HOME/logs by default.
    
35. # export HADOOP_LOG_DIR=${HADOOP_HOME}/logs
    
36. 
    
37. # File naming remote slave hosts.  $HADOOP_HOME/conf/slaves by default.
    
38. # export HADOOP_SLAVES=${HADOOP_HOME}/conf/slaves
    
39. 
    
40. # host:path where hadoop code should be rsync'd from.  Unset by default.
    
41. # export HADOOP_MASTER=master:/home/$USER/src/hadoop
    
42. 
    
43. # Seconds to sleep between slave commands.  Unset by default.  This
    
44. # can be useful in large clusters, where, e.g., slave rsyncs can
    
45. # otherwise arrive faster than the master can service them.
    
46. # export HADOOP_SLAVE_SLEEP=0.1
    
47. 
    
48. # The directory where pid files are stored. /tmp by default.
    
49. # export HADOOP_PID_DIR=/var/hadoop/pids
    
50. 
    
51. # A string representing this instance of hadoop. $USER by default.
    
52. # export HADOOP_IDENT_STRING=$USER
    
53. 
    
54. # The scheduling priority for daemon processes.  See 'man nice'.
    
55. # export HADOOP_NICENESS=10
    
56. 
    
57.   export PATH=$PATH:$HADOOP_HOME/bin

复制代码

最后在配置hadoop环境变量在跟jdk在同一个文件，在jdk环境变量下面增加hadoop环境变量：

1. export HADOOP_HOME=/home/hadoop/hadoop-1.0.4
   
2. export PATH=$PATH:$HADOOP_HOME/bin

复制代码

到这里所有环境配置已经完成：
再一次启动时需要namenode format，命令如下：

1. [hadoop@hadoop bin]$ hadoop namenode -format

复制代码

格式完成后，启动hadoop，使用命令

1. [hadoop@hadoop bin]$ ./start-all.sh

复制代码

使用jps查看hadoop启动的守护进程：

1. [hadoop@hadoop bin]$ jps
   
2. 2734 NameNode
   
3. 2860 DataNode
   
4. 2996 SecondaryNameNode
   
5. 3090 JobTracker
   
6. 3261 TaskTracker

复制代码

hadoop 伪分布式安装已经完成。