Hadoop 实战之MapReduce链接作业之预处理

环境：Vmware 8.0 和Ubuntu11.04

Hadoop 实战之MapReduce链接作业之预处理

第一步：首先创建一个工程命名为HadoopTest.目录结构如下图:

第二步: 在/home/tanglg1987目录下新建一个start.sh脚本文件，每次启动虚拟机都要删除/tmp目录下的全部文件，重新格式化namenode，代码如下：

sudorm-rf/tmp/*
rm-rf/home/tanglg1987/hadoop-0.20.2/logs
hadoopnamenode-format
hadoopdatanode-format
start-all.sh
hadoopfs-mkdirinput
hadoopdfsadmin-safemodeleave

第三步：给start.sh增加执行权限并启动hadoop伪分布式集群，代码如下:

chmod777/home/tanglg1987/start.sh
./start.sh

执行过程如下：

第四步：上传本地文件到hdfs

在/home/tanglg1987目录下新建Customer.txt内容如下：

100tom90
101mary85
102kate60

上传本地文件到hdfs：

hadoopfs-put/home/tanglg1987/ChainMapper.txtinput

第五步：新建一个ChainMapperDemo.java，代码如下：

packagecom.baison.action;
importjava.io.IOException;
importjava.util.*;
importjava.lang.String;
importorg.apache.hadoop.fs.Path;
importorg.apache.hadoop.conf.*;
importorg.apache.hadoop.io.*;
importorg.apache.hadoop.mapred.*;
importorg.apache.hadoop.util.*;
importorg.apache.hadoop.mapred.lib.*;
publicclassChainMapperDemo{
publicstaticclassMap00extendsMapReduceBaseimplements
Mapper<Text,Text,Text,Text>{
publicvoidmap(Textkey,Textvalue,OutputCollectoroutput,
Reporterreporter)throwsIOException{
Textft=newText("100");
if(!key.equals(ft)){
output.collect(key,value);
}
}
}
publicstaticclassMap01extendsMapReduceBaseimplements
Mapper<Text,Text,Text,Text>{
publicvoidmap(Textkey,Textvalue,OutputCollectoroutput,
Reporterreporter)throwsIOException{
Textft=newText("101");
if(!key.equals(ft)){
output.collect(key,value);
}
}
}
publicstaticclassReduceextendsMapReduceBaseimplements
Reducer<Text,Text,Text,Text>{
publicvoidreduce(Textkey,Iteratorvalues,OutputCollectoroutput,
Reporterreporter)throwsIOException{
while(values.hasNext()){
output.collect(key,values.next());
}

}
}
publicstaticvoidmain(String[]args)throwsException{
String[]arg={"hdfs://localhost:9100/user/tanglg1987/input/ChainMapper.txt",
"hdfs://localhost:9100/user/tanglg1987/output"};
JobConfconf=newJobConf(ChainMapperDemo.class);
conf.setJobName("ChainMapperDemo");
conf.setInputFormat(KeyValueTextInputFormat.class);
conf.setOutputFormat(TextOutputFormat.class);
ChainMappercm=newChainMapper();
JobConfmapAConf=newJobConf(false);
cm.addMapper(conf,Map00.class,Text.class,Text.class,Text.class,
Text.class,true,mapAConf);
JobConfmapBConf=newJobConf(false);
cm.addMapper(conf,Map01.class,Text.class,Text.class,Text.class,
Text.class,true,mapBConf);
conf.setReducerClass(Reduce.class);
conf.setOutputKeyClass(Text.class);
conf.setOutputValueClass(Text.class);
FileInputFormat.setInputPaths(conf,newPath(arg[0]));
FileOutputFormat.setOutputPath(conf,newPath(arg[1]));
JobClient.runJob(conf);
}
}

第六步：Run On Hadoop，运行过程如下：

12/10/17 21:05:53 INFO jvm.JvmMetrics: Initializing JVM Metrics with processName=JobTracker, sessionId=
12/10/17 21:05:53 WARN mapred.JobClient: Use GenericOptionsParser for parsing the arguments. Applications should implement Tool for the same.
12/10/17 21:05:53 WARN mapred.JobClient: No job jar file set. User classes may not be found. See JobConf(Class) or JobConf#setJar(String).
12/10/17 21:05:54 INFO mapred.FileInputFormat: Total input paths to process : 1
12/10/17 21:05:54 INFO mapred.JobClient: Running job: job_local_0001
12/10/17 21:05:54 INFO mapred.FileInputFormat: Total input paths to process : 1
12/10/17 21:05:54 INFO mapred.MapTask: numReduceTasks: 1
12/10/17 21:05:54 INFO mapred.MapTask: io.sort.mb = 100
12/10/17 21:05:54 INFO mapred.MapTask: data buffer = 79691776/99614720
12/10/17 21:05:54 INFO mapred.MapTask: record buffer = 262144/327680
12/10/17 21:05:54 INFO mapred.MapTask: Starting flush of map output
12/10/17 21:05:54 INFO mapred.MapTask: Finished spill 0
12/10/17 21:05:54 INFO mapred.TaskRunner: Task:attempt_local_0001_m_000000_0 is done. And is in the process of commiting
12/10/17 21:05:54 INFO mapred.LocalJobRunner: hdfs://localhost:9100/user/tanglg1987/input/ChainMapper.txt:0+35
12/10/17 21:05:54 INFO mapred.TaskRunner: Task 'attempt_local_0001_m_000000_0' done.
12/10/17 21:05:54 INFO mapred.LocalJobRunner:
12/10/17 21:05:54 INFO mapred.Merger: Merging 1 sorted segments
12/10/17 21:05:54 INFO mapred.Merger: Down to the last merge-pass, with 1 segments left of total size: 16 bytes
12/10/17 21:05:54 INFO mapred.LocalJobRunner:
12/10/17 21:05:54 INFO mapred.TaskRunner: Task:attempt_local_0001_r_000000_0 is done. And is in the process of commiting
12/10/17 21:05:54 INFO mapred.LocalJobRunner:
12/10/17 21:05:54 INFO mapred.TaskRunner: Task attempt_local_0001_r_000000_0 is allowed to commit now
12/10/17 21:05:54 INFO mapred.FileOutputCommitter: Saved output of task 'attempt_local_0001_r_000000_0' to hdfs://localhost:9100/user/tanglg1987/output
12/10/17 21:05:54 INFO mapred.LocalJobRunner: reduce > reduce
12/10/17 21:05:54 INFO mapred.TaskRunner: Task 'attempt_local_0001_r_000000_0' done.
12/10/17 21:05:55 INFO mapred.JobClient: map 100% reduce 100%
12/10/17 21:05:55 INFO mapred.JobClient: Job complete: job_local_0001
12/10/17 21:05:55 INFO mapred.JobClient: Counters: 15
12/10/17 21:05:55 INFO mapred.JobClient: FileSystemCounters
12/10/17 21:05:55 INFO mapred.JobClient: FILE_BYTES_READ=36152
12/10/17 21:05:55 INFO mapred.JobClient: HDFS_BYTES_READ=70
12/10/17 21:05:55 INFO mapred.JobClient: FILE_BYTES_WRITTEN=73202
12/10/17 21:05:55 INFO mapred.JobClient: HDFS_BYTES_WRITTEN=12
12/10/17 21:05:55 INFO mapred.JobClient: Map-Reduce Framework
12/10/17 21:05:55 INFO mapred.JobClient: Reduce input groups=1
12/10/17 21:05:55 INFO mapred.JobClient: Combine output records=0
12/10/17 21:05:55 INFO mapred.JobClient: Map input records=3
12/10/17 21:05:55 INFO mapred.JobClient: Reduce shuffle bytes=0
12/10/17 21:05:55 INFO mapred.JobClient: Reduce output records=1
12/10/17 21:05:55 INFO mapred.JobClient: Spilled Records=2
12/10/17 21:05:55 INFO mapred.JobClient: Map output bytes=12
12/10/17 21:05:55 INFO mapred.JobClient: Map input bytes=35
12/10/17 21:05:55 INFO mapred.JobClient: Combine input records=0
12/10/17 21:05:55 INFO mapred.JobClient: Map output records=1
12/10/17 21:05:55 INFO mapred.JobClient: Reduce input records=1

第七步：查看结果集，运行结果如下：

sudorm-rf/tmp/*
rm-rf/home/tanglg1987/hadoop-0.20.2/logs
hadoopnamenode-format
hadoopdatanode-format
start-all.sh
hadoopfs-mkdirinput
hadoopdfsadmin-safemodeleave

第三步：给start.sh增加执行权限并启动hadoop伪分布式集群，代码如下:

chmod777/home/tanglg1987/start.sh
./start.sh