MapReduce之分区
模式描述
分区模式是将记录进行分类，但并不关心记录的顺序
目的
该模式的目的是将数据集中相似的记录分成不同的，更小的数据集
适用场景
适用这一个模式的最主要的要求是：必须提前知道有多少个分区，例如，如果按照天数对周进行分区，那末将会有七个分区
适用场景如下
- 按连续值裁剪分区
- 按类别剪裁分区
- 分片
性能分析
在性能方面这个模式主要关注的是，每个分区的结果数据中是否有类似的数量的记录，有可能一个分区中包含有整个数据集中大约50%的数据，如果直接简单的适用该模式的话，那么该分区对应的数据会全都发送到一个reducer上，此时处理性能便会明显的下降。
问题描述
给定彝族用户的信息，按照最近访问日期中的年份信息对记录进行分区，一年对应一个分区
样例输入
生成数据集的代码如下
import java.io.*;
import java.util.Random;
public class create {
public static String getRandomChar(int length) { //生成随机字符串
Random random = new Random();
StringBuffer buffer = new StringBuffer();
for (int i = 0; i < length; i++) {
buffer.append((char)('a'+random.nextInt(26)));
}
return buffer.toString();
}
public static void main(String[] args) throws IOException{
String path="input/file.txt";
File file=new File(path);
if(!file.exists()){
file.getParentFile().mkdirs();
}
file.createNewFile();
FileWriter fw=new FileWriter(file,true);
BufferedWriter bw=new BufferedWriter(fw);
for(int i=0;i<1000;i++){
int id=(int)(Math.random()*10000+10000);
int lastData=(int)(Math.random()*10+2015);
bw.write("id = "+id+" lastAccessData = "+lastData+" comments = "+getRandomChar(15)+'\n');
}
bw.flush();
bw.close();
fw.close();
}
}
运行结果如下：
mapper阶段任务
mapper获取每条输入记录中的最近访问日期，并将日期中的年份作为键，整条记录作为值输出
mapper阶段编码如下
public static class LastAccessDateMapper extends Mapper<Object,Text, IntWritable,Text>{
private final static SimpleDateFormat frmt=new SimpleDateFormat(
"yyyy");
private IntWritable outkey=new IntWritable();
public void map(Object key,Text value,Context context) throws IOException,InterruptedException{
String line=value.toString();
String strDate=line.substring(line.indexOf("lastAccessData")+17,line.indexOf("lastAccessData")+22);
System.out.println(strDate);
// 通过该方法获取年份
Calendar cal=Calendar.getInstance();
try {
cal.setTime(frmt.parse(strDate));
outkey.set(cal.get(Calendar.YEAR));
} catch (ParseException e) {
e.printStackTrace();
}
System.out.println(outkey);
context.write(outkey,new Text(value));
}
}
partitioner阶段任务
分区器检查mapper输出的每个键/值对，并将确定的键/值对写入相对应的分区，这样在reduce阶段每个已编号的分区数据将被相应的reduce任务复制，在任务初始化阶段配置分区器调用setconf方法，在作业配置阶段，驱动程序负责调用LastAccessDatePartitioner.setMinLastAccessDate方法获取该值，该日期用于减去每个键，来确定每个键将被分配到哪个分区。
public static class LastAccessDatePartitioner extends Partitioner<IntWritable,Text> implements Configurable {
private static final String MIN_LAST_ACCESS_DATE_YEAR="min.last.access.date.year";
private Configuration conf=null;
private int minLastAccessDateYear=0;
public int getPartition(IntWritable key,Text value,int numPartitions){
return key.get()-minLastAccessDateYear;
}
public Configuration getConf(){
return conf;
}
public void setConf(Configuration conf){
this.conf=conf;
minLastAccessDateYear=conf.getInt(MIN_LAST_ACCESS_DATE_YEAR,0);
}
public static void setMinLastAccessDate(Job job,int minLastAccessDateYear){
job.getConfiguration().setInt(MIN_LAST_ACCESS_DATE_YEAR,minLastAccessDateYear);
}
}
reducer阶段
该阶段仅仅只是输出所有的值，所以很简单
reducer阶段编码如下
public static class ValueReducer extends Reducer<IntWritable,Text,Text, NullWritable>{
public void reduce(IntWritable key, Iterable<Text> values,Context context) throws IOException,InterruptedException{
for(Text t:values){
context.write(t,NullWritable.get());
}
}
}
完整代码如下
import org.apache.hadoop.conf.Configurable;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.IntWritable;
import org.apache.hadoop.io.NullWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Job;
import org.apache.hadoop.mapreduce.Mapper;
import org.apache.hadoop.mapreduce.Partitioner;
import org.apache.hadoop.mapreduce.Reducer;
import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;
import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;
import java.io.IOException;
import java.text.ParseException;
import java.text.SimpleDateFormat;
import java.util.Calendar;
public class LastPartition {
public static class LastAccessDateMapper extends Mapper<Object,Text, IntWritable,Text>{
private final static SimpleDateFormat frmt=new SimpleDateFormat(
"yyyy");
private IntWritable outkey=new IntWritable();
public void map(Object key,Text value,Context context) throws IOException,InterruptedException{
String line=value.toString();
String strDate=line.substring(line.indexOf("lastAccessData")+17,line.indexOf("lastAccessData")+22);
System.out.println(strDate);
Calendar cal=Calendar.getInstance();
try {
cal.setTime(frmt.parse(strDate));
outkey.set(cal.get(Calendar.YEAR));
} catch (ParseException e) {
e.printStackTrace();
}
System.out.println(outkey);
context.write(outkey,new Text(value));
}
}
public static class LastAccessDatePartitioner extends Partitioner<IntWritable,Text> implements Configurable {
private static final String MIN_LAST_ACCESS_DATE_YEAR="min.last.access.date.year";
private Configuration conf=null;
private int minLastAccessDateYear=0;
public int getPartition(IntWritable key,Text value,int numPartitions){
return key.get()-minLastAccessDateYear;
}
public Configuration getConf(){
return conf;
}
public void setConf(Configuration conf){
this.conf=conf;
minLastAccessDateYear=conf.getInt(MIN_LAST_ACCESS_DATE_YEAR,0);
}
public static void setMinLastAccessDate(Job job,int minLastAccessDateYear){
job.getConfiguration().setInt(MIN_LAST_ACCESS_DATE_YEAR,minLastAccessDateYear);
}
}
public static class ValueReducer extends Reducer<IntWritable,Text,Text, NullWritable>{
public void reduce(IntWritable key, Iterable<Text> values,Context context) throws IOException,InterruptedException{
for(Text t:values){
context.write(t,NullWritable.get());
}
}
}
public static void main(String[] args) throws Exception{
FileUtil.deleteDir("output");
Configuration configuration=new Configuration();
String[] otherArgs=new String[]{"input/file.txt","output"};
if(otherArgs.length!=2){
System.err.println("参数错误");
System.exit(2);
}
Job job=new Job(configuration,"Layered");
job.setMapperClass(LastAccessDateMapper.class);
job.setReducerClass(ValueReducer.class);
job.setOutputKeyClass(IntWritable.class);
job.setPartitionerClass(LastAccessDatePartitioner.class);
LastAccessDatePartitioner.setMinLastAccessDate(job,2012);
job.setNumReduceTasks(20);
job.setOutputValueClass(Text.class);
FileInputFormat.addInputPath(job,new Path(otherArgs[0]));
FileOutputFormat.setOutputPath(job,new Path(otherArgs[1]));
System.exit(job.waitForCompletion(true)?0:1);
}
}
遇到的问题
在学习过程中遇到如下问题
java.lang.Exception: java.io.IOException: Illegal partition for 2022 (10)
at org.apache.hadoop.mapred.LocalJobRunner$Job.runTasks(LocalJobRunner.java:491)
at org.apache.hadoop.mapred.LocalJobRunner$Job.run(LocalJobRunner.java:551)
Caused by: java.io.IOException: Illegal partition for 2022 (10)
at org.apache.hadoop.mapred.MapTask$MapOutputBuffer.collect(MapTask.java:1089)
at org.apache.hadoop.mapred.MapTask$NewOutputCollector.write(MapTask.java:721)
at org.apache.hadoop.mapreduce.task.TaskInputOutputContextImpl.write(TaskInputOutputContextImpl.java:89)
at org.apache.hadoop.mapreduce.lib.map.WrappedMapper$Context.write(WrappedMapper.java:112)
at LastPartition$LastAccessDateMapper.map(LastPartition.java:35)
at LastPartition$LastAccessDateMapper.map(LastPartition.java:19)
at org.apache.hadoop.mapreduce.Mapper.run(Mapper.java:146)
at org.apache.hadoop.mapred.MapTask.runNewMapper(MapTask.java:793)
at org.apache.hadoop.mapred.MapTask.run(MapTask.java:341)
at org.apache.hadoop.mapred.LocalJobRunner$Job$MapTaskRunnable.run(LocalJobRunner.java:270)
at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
at java.util.concurrent.FutureTask.run(FutureTask.java:266)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
at java.lang.Thread.run(Thread.java:748)
查阅资料后发现是partition和reducetask个数没对上，调整reducetask个数即可
写在最后
求一份适合《MapReduce设计模式》的数据集和一个比较不错的文本处理程序。。。
299
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
