• The following statements adopted From http://trac.nchc.org.tw/cloud/wiki/Hadoop4Win
  • 若您在 Windows 7 底下無法看到 jps 的結果，欲啟動 Hadoop 請執行以下指令:

    taskkill /f /im java.exe
    /opt/hadoop/bin/hadoop-daemon.sh start namenode
    /opt/hadoop/bin/hadoop-daemon.sh start secondarynamenode
    /opt/hadoop/bin/hadoop-daemon.sh start datanode
    /opt/hadoop/bin/hadoop-daemon.sh start jobtracker
    /opt/hadoop/bin/hadoop-daemon.sh start tasktracker
    

• >bin/start-all.sh
  • jps
  • • JobTracker Web UI - 用瀏覽器開啟  http://localhost:50030 的頁面
    • NameNode Web UI - 用瀏覽器開啟  http://localhost:50070 的頁面
  • >bin/hadoop dfsadmin -report
• >bin/stop-all.sh
• test
  • >
• 格式化  NameNode (第一次安裝後做即可)
• >bin/hadoop namenode -format
• 若您在 Windows 7 底下要關閉 Hadoop 的所有服務，請使用以下指令

  /opt/hadoop/bin/hadoop-daemon.sh stop tasktracker
  /opt/hadoop/bin/hadoop-daemon.sh stop datanode
  /opt/hadoop/bin/hadoop-daemon.sh stop secondarynamenode
  /opt/hadoop/bin/hadoop-daemon.sh stop jobtracker
  /opt/hadoop/bin/hadoop-daemon.sh stop namenode

HDFS 相關操作範例

• >bin/hadoop namenode -format

•  >bin/hadoop fs -mkdir /user/root/tmp

• >bin/hadoop fs -ls /user/root/tmp

• >bin/hadoop fs -put conf/* /user/root/tmp

• >bin/hadoop fs -ls /user/root/tmp

• >bin/hadoop fs -cat /user/root/tmp/core-site.xml

• >bin/hadoop fs -get /user/root/tmp/core-site.xml /opt/hadoop/

• >bin/hadoop fs -rm /user/root/tmp/core-site.xml

• >bin/hadoop fs -rmr /user/root/tmp

• NameNode http://localhost:50070/

•  

MapReduce 相關操作範例

• >bin/hadoop job -list all

• >bin/hadoop job -status job_??????_????

• >bin/hadoop job -history /user/root/output

• >bin/hadoop jar hadoop-0.20.2-examples.jar grep input output 'config[a-z.]+'

• JobTracker http://localhost:50030/jobtracker.jsp

HBase 相關操作範例

• (hadoop4win) start-hbase

• >cd /opt/hbase

• >bin/hbase shell

• hbase(main):001:0>create 'scores'.'studentid', 'course'

• hbase(main):001:0>list

• hbase(main):001:0>describe 'scores'

• hbase(main):001:0>put 'scores', 'jdwang', 'studentid:', '1'

• hbase(main):001:0>put 'scores','jdwang','course:math','80'

• hbase(main):001:0>put 'scores','jdwang','course:history','85'

• hbase(main):001:0>put 'scores','lucy','studentid,'2'

• hbase(main):001:0>put 'scores','lucy','course:math','75'

• hbase(main):001:0>put 'scores','lucy','course:history','90'

• hbase(main):001:0>get 'scores','jdwang'

• hbase(main):001:0>get 'scores','lucy'

• hbase(main):001:0>scan 'scores'

• hbase(main):001:0>scan 'scores',{COLUMNS=>'course:'}

• hbase(main):001:0>scan 'scores',{COLUMNS=>['studnets','course:']}

• hbase(main):001:0>disable 'scores'

• hbase(main):001:0>drop 'scores'

• HMaster http://localhost:60010/

• HRegionServer http://localhost:60030/

• ZooKeeper http://localhost:60010/

cd $HADOOP_HOME

=> cd opt/hadoop
$  mkdir inputwc
$  bin/hadoop dfs -mkdir input
$  echo "This is a test" > inputwc/input1
$  echo "That is also a test" > inputwc/input2
$  bin/hadoop dfs -put inputwc input
$  bin/hadoop dfs -ls input
http://trac.nchc.org.tw/cloud/attachment/wiki/jazz/Hadoop_Lab6/WordCount.java?format=raw
wordcount.java => wordcount_jdwang1.java
line10=>public class WordCount_jdwang1 {
line37=>JobConf conf = new JobConf(WordCount_jdwang1.class);

$ mkdir MyJava
$  javac -classpath hadoop-0.20.2-core.jar -d MyJava WordCount_jdwang.java
$  jar -cvf wordcount_jdwang1.jar -C Myjava .
$  bin/hadoop jar wordcount_jdwang.jar WordCount_jdwang1 input/ output/

=>/opt/hadoop/bin/hadoop fs -rmr output （要先清掉舊資料才可以得到新的結果）
$ bin/hadoop dfs -ls output
$ bin/hadoop dfs -cat output/part-00000
$ bin/hadoop dfs -copyToLocal output/part-00000 output/part-00000_jdwang.txt

=================================================================

http://trac.nchc.org.tw/cloud/raw-attachment/wiki/jazz/Hadoop_Lab6/WordCount2.java
line11=>public class WordCount2_jdwang2 extends Configured implements Tool {
line86=>JobConf conf = new JobConf(getConf(), WordCount2_jdwang2.class);
line117=>int res = ToolRunner.run(new Configuration(), new WordCount2_jdwang2(), args);
$ echo “\.” > pattern.txt
$ echo “\,” >> pattern.txt
$ mkdir MyJava2
$ javac -classpath hadoop-0.20.2-core.jar -d MyJava2 WordCount2_jdwang.java
$ jar -cvf wordcount2_jdwang2.jar -C MyJava2 .
$ bin/hadoop jar wordcount2.jar WordCount2_jdwang input output2 -skip pattern.txt
$ bin/hadoop dfs -cat output2/part-00000
$ bin/hadoop dfs -copyFromLocal output2 output2

http://localhost:50030/jobtracker.jsp

http://localhost:50070/dfshealth.jsp