Hadoop基础-MapReduce的Partitioner用法案例

　　　　　　　　　　　　　　　　　　　　　　Hadoop基础-MapReduce的Partitioner用法案例

　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　作者：尹正杰

版权声明：原创作品，谢绝转载！否则将追究法律责任。

 

 

 

一.Partitioner关键代码剖析

1>.返回的分区号

2>.partitioner默认是通过hash方法实现的

 　　返回的是一个int类型的数组：

3>.HashPartitioner

　　接下来咱们就看看Partition在MapReduce的作用是什么吧。

 

二.Partitioner在MapReduce的位置

1>.什么是Partitioner

　　通过查看Partitioner的源码估计你也发现了Partitioner跟hash有关系，那么它到底是什么呢？能帮我们做什么事情呢？说白了它就是帮我们实现分发Key和value的一个过程，它负责将同一个key发给同一个Reduce。还记得我们之前说的Combiner吗？Combiner相当于Map端的Reduce，用于减少网络间分发。说直白点就是将key进行折叠的一个操作，将一个key的初始值为1方给Reduce端100万次，和将key进行折叠后形成key的初始值为100万，在发送给Reduce可以街上带宽资源，还可以减少网络带宽。而这个分发过程就是Partitioner程序完成的，当然我们是不定义Partitioner也不会报错。

2>.Partitioner在MapReduce的位置

 　　接下来我们大致看一下Partitioner在MapReduce的大致位置，如下：

 

　　接下来我们就一起体验一下设置Partitioner和不设置Partitioner的明显区别。

 

三.未定义Partitioner的情况

1>.测试数据文件内容（partitioner.txt）

yinzhengjie    1
golang    2
python    3
shell    4
java    5
linux    6
vbs    7
c++    8
css    9
html    10
javascript    11
尹正杰    12
yinzhengjie    13
golang    14
python    15
shell    16
java    17
linux    18
vbs    19
c++    20
css    21
html    22
javascript    23
尹正杰    24

2>.KVMapper.java 文件内容

 1 /*
 2 @author :yinzhengjie
 3 Blog:http://www.cnblogs.com/yinzhengjie/tag/Hadoop%E8%BF%9B%E9%98%B6%E4%B9%8B%E8%B7%AF/
 4 EMAIL:[email protected]
 5 */
 6 package cn.org.yinzhengjie.mapreduce.partition;
 7 
 8 import org.apache.hadoop.io.IntWritable;
 9 import org.apache.hadoop.io.Text;
10 import org.apache.hadoop.mapreduce.Mapper;
11 
12 import java.io.IOException;
13 
14 public class KVMapper extends Mapper {
15     @Override
16     protected void map(Text key, Text value, Context context) throws IOException, InterruptedException {
17             //将value转换成int类型
18             int val = Integer.parseInt(value.toString());
19             context.write(key,new IntWritable(val));
20     }
21 }

3>.KVReduce.java 文件内容

 1 /*
 2 @author :yinzhengjie
 3 Blog:http://www.cnblogs.com/yinzhengjie/tag/Hadoop%E8%BF%9B%E9%98%B6%E4%B9%8B%E8%B7%AF/
 4 EMAIL:[email protected]
 5 */
 6 package cn.org.yinzhengjie.mapreduce.partition;
 7 
 8 import org.apache.hadoop.io.IntWritable;
 9 import org.apache.hadoop.io.Text;
10 import org.apache.hadoop.mapreduce.Reducer;
11 
12 import java.io.IOException;
13 
14 public class KVReduce extends Reducer {
15     @Override
16     protected void reduce(Text key, Iterable values, Context context) throws IOException, InterruptedException {
17         int count = 0;
18         for (IntWritable value : values) {
19             count += value.get();
20         }
21         context.write(key,new IntWritable(count));
22     }
23 }

4>.KVApp.java 文件内容

 1 /*
 2 @author :yinzhengjie
 3 Blog:http://www.cnblogs.com/yinzhengjie/tag/Hadoop%E8%BF%9B%E9%98%B6%E4%B9%8B%E8%B7%AF/
 4 EMAIL:[email protected]
 5 */
 6 package cn.org.yinzhengjie.mapreduce.partition;
 7 
 8 import org.apache.hadoop.conf.Configuration;
 9 import org.apache.hadoop.fs.FileSystem;
10 import org.apache.hadoop.fs.Path;
11 import org.apache.hadoop.io.IntWritable;
12 import org.apache.hadoop.io.Text;
13 import org.apache.hadoop.mapreduce.Job;
14 import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;
15 import org.apache.hadoop.mapreduce.lib.input.KeyValueTextInputFormat;
16 import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;
17 
18 public class KVApp {
19     public static void main(String[] args) throws Exception {
20         //实例化一个Configuration，它会自动去加载本地的core-site.xml配置文件的fs.defaultFS属性。(该文件放在项目的resources目录即可。)
21         Configuration conf = new Configuration();
22         //将hdfs写入的路径定义在本地，需要修改默认为文件系统，这样就可以覆盖到之前在core-site.xml配置文件读取到的数据。
23         conf.set("fs.defaultFS","file:///");
24         //创建一个任务对象job，别忘记把conf穿进去哟！
25         Job job = Job.getInstance(conf);
26         //给任务起个名字
27         job.setJobName("WordCount");
28         //设置输入格式以K-V的类型传入，这样K的类型就是Mapper输入端的key，而V的类型就是Mapper输入端的value
29         job.setInputFormatClass(KeyValueTextInputFormat.class);
30         //指定main函数所在的类，也就是当前所在的类名
31         job.setJarByClass(KVApp.class);
32         //指定map的类名,这里指定咱们自定义的map程序即可
33         job.setMapperClass(KVMapper.class);
34         //指定reduce的类名，这里指定咱们自定义的reduce程序即可
35         job.setReducerClass(KVReduce.class);
36         //设置输出key的数据类型
37         job.setOutputKeyClass(Text.class);
38         //设置输出value的数据类型
39         job.setOutputValueClass(IntWritable.class);
40         //设置输入路径，需要传递两个参数，即任务对象(job)以及输入路径
41         FileInputFormat.addInputPath(job,new Path("D:\\10.Java\\IDE\\yhinzhengjieData\\MyHadoop\\Partitioner\\partitioner.txt"));
42         //初始化HDFS文件系统，此时我们需要把读取到的fs.defaultFS属性传给fs对象。我的目的是调用该对象的delete方法，删除已经存在的文件夹
43         FileSystem fs = FileSystem.get(conf);
44         //通过fs的delete方法可以删除文件，第一个参数指的是删除文件对象，第二参数是指递归删除，一般用作删除目录
45         Path outPath = new Path("D:\\10.Java\\IDE\\yhinzhengjieData\\MyHadoop\\Partitioner\\out");
46         if (fs.exists(outPath)){
47             fs.delete(outPath,true);
48         }
49         //设置输出路径，需要传递两个参数，即任务对象(job)以及输出路径
50         FileOutputFormat.setOutputPath(job,outPath);
51         //Reduce的个数，咱们是可以自己设置的
52         job.setNumReduceTasks(2);
53         //等待任务执行结束,将里面的值设置为true。
54         job.waitForCompletion(true);
55     }
56 }

WARNING: An illegal reflective access operation has occurred
WARNING: Illegal reflective access by org.apache.hadoop.security.authentication.util.KerberosUtil (file:/C:/Users/Administrator/.m2/repository/org/apache/hadoop/hadoop-auth/2.7.3/hadoop-auth-2.7.3.jar) to method sun.security.krb5.Config.getInstance()
WARNING: Please consider reporting this to the maintainers of org.apache.hadoop.security.authentication.util.KerberosUtil
WARNING: Use --illegal-access=warn to enable warnings of further illegal reflective access operations
WARNING: All illegal access operations will be denied in a future release
18/06/18 08:51:19 INFO Configuration.deprecation: session.id is deprecated. Instead, use dfs.metrics.session-id
18/06/18 08:51:19 INFO jvm.JvmMetrics: Initializing JVM Metrics with processName=JobTracker, sessionId=
18/06/18 08:51:19 WARN mapreduce.JobResourceUploader: Hadoop command-line option parsing not performed. Implement the Tool interface and execute your application with ToolRunner to remedy this.
18/06/18 08:51:19 WARN mapreduce.JobResourceUploader: No job jar file set.  User classes may not be found. See Job or Job#setJar(String).
18/06/18 08:51:19 INFO input.FileInputFormat: Total input paths to process : 1
18/06/18 08:51:19 INFO mapreduce.JobSubmitter: number of splits:1
18/06/18 08:51:19 INFO mapreduce.JobSubmitter: Submitting tokens for job: job_local1080098995_0001
18/06/18 08:51:19 INFO mapreduce.Job: The url to track the job: http://localhost:8080/
18/06/18 08:51:19 INFO mapreduce.Job: Running job: job_local1080098995_0001
18/06/18 08:51:19 INFO mapred.LocalJobRunner: OutputCommitter set in config null
18/06/18 08:51:19 INFO output.FileOutputCommitter: File Output Committer Algorithm version is 1
18/06/18 08:51:19 INFO mapred.LocalJobRunner: OutputCommitter is org.apache.hadoop.mapreduce.lib.output.FileOutputCommitter
18/06/18 08:51:19 INFO mapred.LocalJobRunner: Waiting for map tasks
18/06/18 08:51:19 INFO mapred.LocalJobRunner: Starting task: attempt_local1080098995_0001_m_000000_0
18/06/18 08:51:19 INFO output.FileOutputCommitter: File Output Committer Algorithm version is 1
18/06/18 08:51:19 INFO util.ProcfsBasedProcessTree: ProcfsBasedProcessTree currently is supported only on Linux.
18/06/18 08:51:19 INFO mapred.Task:  Using ResourceCalculatorProcessTree : org.apache.hadoop.yarn.util.WindowsBasedProcessTree@59eafdde
18/06/18 08:51:19 INFO mapred.MapTask: Processing split: file:/D:/10.Java/IDE/yhinzhengjieData/MyHadoop/Partitioner/partitioner.txt:0+241
18/06/18 08:51:19 INFO mapred.MapTask: (EQUATOR) 0 kvi 26214396(104857584)
18/06/18 08:51:19 INFO mapred.MapTask: mapreduce.task.io.sort.mb: 100
18/06/18 08:51:19 INFO mapred.MapTask: soft limit at 83886080
18/06/18 08:51:19 INFO mapred.MapTask: bufstart = 0; bufvoid = 104857600
18/06/18 08:51:19 INFO mapred.MapTask: kvstart = 26214396; length = 6553600
18/06/18 08:51:19 INFO mapred.MapTask: Map output collector class = org.apache.hadoop.mapred.MapTask$MapOutputBuffer
18/06/18 08:51:19 INFO mapred.LocalJobRunner: 
18/06/18 08:51:19 INFO mapred.MapTask: Starting flush of map output
18/06/18 08:51:19 INFO mapred.MapTask: Spilling map output
18/06/18 08:51:19 INFO mapred.MapTask: bufstart = 0; bufend = 252; bufvoid = 104857600
18/06/18 08:51:19 INFO mapred.MapTask: kvstart = 26214396(104857584); kvend = 26214304(104857216); length = 93/6553600
18/06/18 08:51:19 INFO mapred.MapTask: Finished spill 0
18/06/18 08:51:19 INFO mapred.Task: Task:attempt_local1080098995_0001_m_000000_0 is done. And is in the process of committing
18/06/18 08:51:19 INFO mapred.LocalJobRunner: file:/D:/10.Java/IDE/yhinzhengjieData/MyHadoop/Partitioner/partitioner.txt:0+241
18/06/18 08:51:19 INFO mapred.Task: Task 'attempt_local1080098995_0001_m_000000_0' done.
18/06/18 08:51:19 INFO mapred.LocalJobRunner: Finishing task: attempt_local1080098995_0001_m_000000_0
18/06/18 08:51:19 INFO mapred.LocalJobRunner: map task executor complete.
18/06/18 08:51:19 INFO mapred.LocalJobRunner: Waiting for reduce tasks
18/06/18 08:51:19 INFO mapred.LocalJobRunner: Starting task: attempt_local1080098995_0001_r_000000_0
18/06/18 08:51:19 INFO output.FileOutputCommitter: File Output Committer Algorithm version is 1
18/06/18 08:51:19 INFO util.ProcfsBasedProcessTree: ProcfsBasedProcessTree currently is supported only on Linux.
18/06/18 08:51:19 INFO mapred.Task:  Using ResourceCalculatorProcessTree : org.apache.hadoop.yarn.util.WindowsBasedProcessTree@5d916cac
18/06/18 08:51:19 INFO mapred.ReduceTask: Using ShuffleConsumerPlugin: org.apache.hadoop.mapreduce.task.reduce.Shuffle@6ef6c4ad
18/06/18 08:51:19 INFO reduce.MergeManagerImpl: MergerManager: memoryLimit=2996200960, maxSingleShuffleLimit=749050240, mergeThreshold=1977492736, ioSortFactor=10, memToMemMergeOutputsThreshold=10
18/06/18 08:51:19 INFO reduce.EventFetcher: attempt_local1080098995_0001_r_000000_0 Thread started: EventFetcher for fetching Map Completion Events
18/06/18 08:51:19 INFO reduce.LocalFetcher: localfetcher#1 about to shuffle output of map attempt_local1080098995_0001_m_000000_0 decomp: 144 len: 148 to MEMORY
18/06/18 08:51:19 INFO reduce.InMemoryMapOutput: Read 144 bytes from map-output for attempt_local1080098995_0001_m_000000_0
18/06/18 08:51:19 INFO reduce.MergeManagerImpl: closeInMemoryFile -> map-output of size: 144, inMemoryMapOutputs.size() -> 1, commitMemory -> 0, usedMemory ->144
18/06/18 08:51:19 INFO reduce.EventFetcher: EventFetcher is interrupted.. Returning
18/06/18 08:51:19 INFO mapred.LocalJobRunner: 1 / 1 copied.
18/06/18 08:51:19 INFO reduce.MergeManagerImpl: finalMerge called with 1 in-memory map-outputs and 0 on-disk map-outputs
18/06/18 08:51:19 INFO mapred.Merger: Merging 1 sorted segments
18/06/18 08:51:19 INFO mapred.Merger: Down to the last merge-pass, with 1 segments left of total size: 138 bytes
18/06/18 08:51:19 INFO reduce.MergeManagerImpl: Merged 1 segments, 144 bytes to disk to satisfy reduce memory limit
18/06/18 08:51:19 INFO reduce.MergeManagerImpl: Merging 1 files, 148 bytes from disk
18/06/18 08:51:19 INFO reduce.MergeManagerImpl: Merging 0 segments, 0 bytes from memory into reduce
18/06/18 08:51:19 INFO mapred.Merger: Merging 1 sorted segments
18/06/18 08:51:19 INFO mapred.Merger: Down to the last merge-pass, with 1 segments left of total size: 138 bytes
18/06/18 08:51:19 INFO mapred.LocalJobRunner: 1 / 1 copied.
18/06/18 08:51:19 INFO Configuration.deprecation: mapred.skip.on is deprecated. Instead, use mapreduce.job.skiprecords
18/06/18 08:51:19 INFO mapred.Task: Task:attempt_local1080098995_0001_r_000000_0 is done. And is in the process of committing
18/06/18 08:51:19 INFO mapred.LocalJobRunner: 1 / 1 copied.
18/06/18 08:51:19 INFO mapred.Task: Task attempt_local1080098995_0001_r_000000_0 is allowed to commit now
18/06/18 08:51:19 INFO output.FileOutputCommitter: Saved output of task 'attempt_local1080098995_0001_r_000000_0' to file:/D:/10.Java/IDE/yhinzhengjieData/MyHadoop/Partitioner/out/_temporary/0/task_local1080098995_0001_r_000000
18/06/18 08:51:19 INFO mapred.LocalJobRunner: reduce > reduce
18/06/18 08:51:19 INFO mapred.Task: Task 'attempt_local1080098995_0001_r_000000_0' done.
18/06/18 08:51:19 INFO mapred.LocalJobRunner: Finishing task: attempt_local1080098995_0001_r_000000_0
18/06/18 08:51:19 INFO mapred.LocalJobRunner: Starting task: attempt_local1080098995_0001_r_000001_0
18/06/18 08:51:19 INFO output.FileOutputCommitter: File Output Committer Algorithm version is 1
18/06/18 08:51:19 INFO util.ProcfsBasedProcessTree: ProcfsBasedProcessTree currently is supported only on Linux.
18/06/18 08:51:19 INFO mapred.Task:  Using ResourceCalculatorProcessTree : org.apache.hadoop.yarn.util.WindowsBasedProcessTree@378ab11d
18/06/18 08:51:19 INFO mapred.ReduceTask: Using ShuffleConsumerPlugin: org.apache.hadoop.mapreduce.task.reduce.Shuffle@10246d23
18/06/18 08:51:19 INFO reduce.MergeManagerImpl: MergerManager: memoryLimit=2996200960, maxSingleShuffleLimit=749050240, mergeThreshold=1977492736, ioSortFactor=10, memToMemMergeOutputsThreshold=10
18/06/18 08:51:19 INFO reduce.EventFetcher: attempt_local1080098995_0001_r_000001_0 Thread started: EventFetcher for fetching Map Completion Events
18/06/18 08:51:19 INFO reduce.LocalFetcher: localfetcher#2 about to shuffle output of map attempt_local1080098995_0001_m_000000_0 decomp: 160 len: 164 to MEMORY
18/06/18 08:51:19 INFO reduce.InMemoryMapOutput: Read 160 bytes from map-output for attempt_local1080098995_0001_m_000000_0
18/06/18 08:51:19 INFO reduce.MergeManagerImpl: closeInMemoryFile -> map-output of size: 160, inMemoryMapOutputs.size() -> 1, commitMemory -> 0, usedMemory ->160
18/06/18 08:51:19 INFO reduce.EventFetcher: EventFetcher is interrupted.. Returning
18/06/18 08:51:19 INFO mapred.LocalJobRunner: 1 / 1 copied.
18/06/18 08:51:19 INFO reduce.MergeManagerImpl: finalMerge called with 1 in-memory map-outputs and 0 on-disk map-outputs
18/06/18 08:51:19 INFO mapred.Merger: Merging 1 sorted segments
18/06/18 08:51:19 INFO mapred.Merger: Down to the last merge-pass, with 1 segments left of total size: 151 bytes
18/06/18 08:51:19 INFO reduce.MergeManagerImpl: Merged 1 segments, 160 bytes to disk to satisfy reduce memory limit
18/06/18 08:51:19 INFO reduce.MergeManagerImpl: Merging 1 files, 164 bytes from disk
18/06/18 08:51:19 INFO reduce.MergeManagerImpl: Merging 0 segments, 0 bytes from memory into reduce
18/06/18 08:51:19 INFO mapred.Merger: Merging 1 sorted segments
18/06/18 08:51:19 INFO mapred.Merger: Down to the last merge-pass, with 1 segments left of total size: 151 bytes
18/06/18 08:51:19 INFO mapred.LocalJobRunner: 1 / 1 copied.
18/06/18 08:51:19 INFO mapred.Task: Task:attempt_local1080098995_0001_r_000001_0 is done. And is in the process of committing
18/06/18 08:51:19 INFO mapred.LocalJobRunner: 1 / 1 copied.
18/06/18 08:51:19 INFO mapred.Task: Task attempt_local1080098995_0001_r_000001_0 is allowed to commit now
18/06/18 08:51:19 INFO output.FileOutputCommitter: Saved output of task 'attempt_local1080098995_0001_r_000001_0' to file:/D:/10.Java/IDE/yhinzhengjieData/MyHadoop/Partitioner/out/_temporary/0/task_local1080098995_0001_r_000001
18/06/18 08:51:19 INFO mapred.LocalJobRunner: reduce > reduce
18/06/18 08:51:19 INFO mapred.Task: Task 'attempt_local1080098995_0001_r_000001_0' done.
18/06/18 08:51:19 INFO mapred.LocalJobRunner: Finishing task: attempt_local1080098995_0001_r_000001_0
18/06/18 08:51:19 INFO mapred.LocalJobRunner: reduce task executor complete.
18/06/18 08:51:20 INFO mapreduce.Job: Job job_local1080098995_0001 running in uber mode : false
18/06/18 08:51:20 INFO mapreduce.Job:  map 100% reduce 100%
18/06/18 08:51:20 INFO mapreduce.Job: Job job_local1080098995_0001 completed successfully
18/06/18 08:51:20 INFO mapreduce.Job: Counters: 30
    File System Counters
        FILE: Number of bytes read=2721
        FILE: Number of bytes written=886698
        FILE: Number of read operations=0
        FILE: Number of large read operations=0
        FILE: Number of write operations=0
    Map-Reduce Framework
        Map input records=24
        Map output records=24
        Map output bytes=252
        Map output materialized bytes=312
        Input split bytes=139
        Combine input records=0
        Combine output records=0
        Reduce input groups=12
        Reduce shuffle bytes=312
        Reduce input records=24
        Reduce output records=12
        Spilled Records=48
        Shuffled Maps =2
        Failed Shuffles=0
        Merged Map outputs=2
        GC time elapsed (ms)=0
        Total committed heap usage (bytes)=805306368
    Shuffle Errors
        BAD_ID=0
        CONNECTION=0
        IO_ERROR=0
        WRONG_LENGTH=0
        WRONG_MAP=0
        WRONG_REDUCE=0
    File Input Format Counters 
        Bytes Read=241
    File Output Format Counters 
        Bytes Written=138

以上代码输出结果

5>.查看生成的文件内容

 

四.定义Partitioner的情况

1>.Mapper和Reduce端代码不变

 1 /*
 2 @author :yinzhengjie
 3 Blog:http://www.cnblogs.com/yinzhengjie/tag/Hadoop%E8%BF%9B%E9%98%B6%E4%B9%8B%E8%B7%AF/
 4 EMAIL:[email protected]
 5 */
 6 package cn.org.yinzhengjie.mapreduce.partition;
 7 
 8 import org.apache.hadoop.io.IntWritable;
 9 import org.apache.hadoop.io.Text;
10 import org.apache.hadoop.mapreduce.Mapper;
11 
12 import java.io.IOException;
13 
14 public class KVMapper extends Mapper {
15     @Override
16     protected void map(Text key, Text value, Context context) throws IOException, InterruptedException {
17             //将value转换成int类型
18             int val = Integer.parseInt(value.toString());
19             context.write(key,new IntWritable(val));
20     }
21 }

KVMapper.java 文件内容

 1 /*
 2 @author :yinzhengjie
 3 Blog:http://www.cnblogs.com/yinzhengjie/tag/Hadoop%E8%BF%9B%E9%98%B6%E4%B9%8B%E8%B7%AF/
 4 EMAIL:[email protected]
 5 */
 6 package cn.org.yinzhengjie.mapreduce.partition;
 7 
 8 import org.apache.hadoop.io.IntWritable;
 9 import org.apache.hadoop.io.Text;
10 import org.apache.hadoop.mapreduce.Reducer;
11 
12 import java.io.IOException;
13 
14 public class KVReduce extends Reducer {
15     @Override
16     protected void reduce(Text key, Iterable values, Context context) throws IOException, InterruptedException {
17         int count = 0;
18         for (IntWritable value : values) {
19             count += value.get();
20         }
21         context.write(key,new IntWritable(count));
22     }
23 }

KVReduce.java 文件内容

2>.Partitioner端代码

 1 /*
 2 @author :yinzhengjie
 3 Blog:http://www.cnblogs.com/yinzhengjie/tag/Hadoop%E8%BF%9B%E9%98%B6%E4%B9%8B%E8%B7%AF/
 4 EMAIL:[email protected]
 5 */
 6 package cn.org.yinzhengjie.mapreduce.partition;
 7 
 8 import org.apache.hadoop.io.IntWritable;
 9 import org.apache.hadoop.io.Text;
10 import org.apache.hadoop.mapreduce.Partitioner;
11 
12 public class Partition extends Partitioner {
13     @Override
14     public int getPartition(Text text, IntWritable intWritable, int numPartitions) {
15         //此处使用了一个取巧的方式，如果字符串不能转换成数字，说明该字符串是数字类型，就会被分到0号分区，反之就会分到1号分区。
16         try {
17             Integer.parseInt(text.toString());
18             return 0;
19         } catch (Exception e) {
20             return 1;
21         }
22     }
23 }

3>.KVApp.java 端代码

 1 /*
 2 @author :yinzhengjie
 3 Blog:http://www.cnblogs.com/yinzhengjie/tag/Hadoop%E8%BF%9B%E9%98%B6%E4%B9%8B%E8%B7%AF/
 4 EMAIL:[email protected]
 5 */
 6 package cn.org.yinzhengjie.mapreduce.partition;
 7 
 8 import org.apache.hadoop.conf.Configuration;
 9 import org.apache.hadoop.fs.FileSystem;
10 import org.apache.hadoop.fs.Path;
11 import org.apache.hadoop.io.IntWritable;
12 import org.apache.hadoop.io.Text;
13 import org.apache.hadoop.mapreduce.Job;
14 import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;
15 import org.apache.hadoop.mapreduce.lib.input.KeyValueTextInputFormat;
16 import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;
17 
18 public class KVApp {
19     public static void main(String[] args) throws Exception {
20         //实例化一个Configuration，它会自动去加载本地的core-site.xml配置文件的fs.defaultFS属性。(该文件放在项目的resources目录即可。)
21         Configuration conf = new Configuration();
22         //将hdfs写入的路径定义在本地，需要修改默认为文件系统，这样就可以覆盖到之前在core-site.xml配置文件读取到的数据。
23         conf.set("fs.defaultFS","file:///");
24         //创建一个任务对象job，别忘记把conf穿进去哟！
25         Job job = Job.getInstance(conf);
26         //给任务起个名字
27         job.setJobName("WordCount");
28         //设置输入格式以K-V的类型传入，这样K的类型就是Mapper输入端的key，而V的类型就是Mapper输入端的value
29         job.setInputFormatClass(KeyValueTextInputFormat.class);
30         //指定main函数所在的类，也就是当前所在的类名
31         job.setJarByClass(KVApp.class);
32         //指定map的类名,这里指定咱们自定义的map程序即可
33         job.setMapperClass(KVMapper.class);
34         //指定reduce的类名，这里指定咱们自定义的reduce程序即可
35         job.setReducerClass(KVReduce.class);
36         //指定Partitioner的类名，这里指定咱们自定义的Partition程序即可
37         job.setPartitionerClass(Partition.class);
38         //设置输出key的数据类型
39         job.setOutputKeyClass(Text.class);
40         //设置输出value的数据类型
41         job.setOutputValueClass(IntWritable.class);
42         //设置输入路径，需要传递两个参数，即任务对象(job)以及输入路径
43         FileInputFormat.addInputPath(job,new Path("D:\\10.Java\\IDE\\yhinzhengjieData\\MyHadoop\\Partitioner\\partitioner.txt"));
44         //初始化HDFS文件系统，此时我们需要把读取到的fs.defaultFS属性传给fs对象。我的目的是调用该对象的delete方法，删除已经存在的文件夹
45         FileSystem fs = FileSystem.get(conf);
46         //通过fs的delete方法可以删除文件，第一个参数指的是删除文件对象，第二参数是指递归删除，一般用作删除目录
47         Path outPath = new Path("D:\\10.Java\\IDE\\yhinzhengjieData\\MyHadoop\\Partitioner\\out");
48         if (fs.exists(outPath)){
49             fs.delete(outPath,true);
50         }
51         //设置输出路径，需要传递两个参数，即任务对象(job)以及输出路径
52         FileOutputFormat.setOutputPath(job,outPath);
53         //Reduce的个数，咱们是可以自己设置的
54         job.setNumReduceTasks(2);
55         //等待任务执行结束,将里面的值设置为true。
56         job.waitForCompletion(true);
57     }
58 }

4>.测试结果如下：