1、顺序组合式
顺序组合式就是按照指定顺序执行任务如:mapreduce1 --> mapreduce2 --> mapreduce3
即:mapreduce1的输出是mapreduce2的输入,mapreduce2的输出式mapreduce3的输入
代码片段如下:
String inPath1 = "hdfs://hadoop0:9000/user/root/3D/";
String outPath1 = "hdfs://hadoop0:9000/user/root/3DZout/";
String outPath2 = "hdfs://hadoop0:9000/user/root/3DZout2/";
String outPath3 = "hdfs://hadoop0:9000/user/root/3DZout3/";
// job1配置
Job job1 = Job.getInstance(conf);
job1.setJarByClass(Mode.class);
job1.setMapperClass(Map1.class);
job1.setReducerClass(Reduce1.class);
job1.setMapOutputKeyClass(Text.class);
job1.setMapOutputValueClass(IntWritable.class);
job1.setOutputKeyClass(Text.class);
job1.setOutputValueClass(IntWritable.class);
FileInputFormat.addInputPath(job1, new Path(inPath1));
FileOutputFormat.setOutputPath(job1, new Path(outPath1));
job1.waitForCompletion(true);
// job2配置
Job job2 = Job.getInstance(conf);
job2.setJarByClass(Mode.class);
job2.setMapperClass(Map2.class);
job2.setReducerClass(Reduce2.class);
job2.setMapOutputKeyClass(Text.class);
job2.setMapOutputValueClass(IntWritable.class);
job2.setOutputKeyClass(Text.class);
job2.setOutputValueClass(IntWritable.class);
FileInputFormat.addInputPath(job2, new Path(inPath1));
FileOutputFormat.setOutputPath(job2, new Path(outPath2));
job2.waitForCompletion(true);
// job3配置
Job job3 = Job.getInstance(conf);
job3.setJarByClass(Mode.class);
job3.setMapperClass(Map3.class);
job3.setReducerClass(Reduce3.class);
job3.setMapOutputKeyClass(Text.class);
job3.setMapOutputValueClass(IntWritable.class);
job3.setOutputKeyClass(Text.class);
job3.setOutputValueClass(IntWritable.class);
FileInputFormat.addInputPath(job3, new Path(outPath2));
FileOutputFormat.setOutputPath(job3, new Path(outPath3));
job3.waitForCompletion(true);
子任务作业配置代码运行后,将按照顺序逐个执行每个子任务作业。由于后一个子任务需要使用前一个子任务的输出数据,因此,每一个子任务
都需要等前一个子任务执行执行完毕后才允许执行,这是通过job.waitForCompletion(true)方法加以保证的。
2、迭代组合式
迭代也可以理解为for循环或while循环,当满足某些条件时,循环结束
mapreduce的迭代算法正在研究中,后续提供完整源码....
代码如下:
3、复杂的依赖组合式
处理复杂的要求的时候,有时候一个mapreduce程序完成不了,往往需要多个mapreduce程序 这个时候就牵扯到各个任务之间的依赖关系,
所谓依赖就是一个M/R job的处理结果是另外一个M/R的输入,以此类推,
这里的顺序是 job1 和 job2 单独执行, job3依赖job1和job2执行后的结果
代码如下:
package com.hadoop.mapreduce;
import java.io.IOException;
import java.util.StringTokenizer;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.FileSystem;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.IntWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Job;
import org.apache.hadoop.mapreduce.Mapper;
import org.apache.hadoop.mapreduce.Reducer;
import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;
import org.apache.hadoop.mapreduce.lib.jobcontrol.ControlledJob;
import org.apache.hadoop.mapreduce.lib.jobcontrol.JobControl;
import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;
import org.apache.hadoop.util.GenericOptionsParser;
public class Mode {
// 第一个Job
public static class Map1 extends Mapper<Object, Text, Text, IntWritable>{
Text word = new Text();
@Override
protected void map(Object key, Text value,Context context)
throws IOException, InterruptedException {
StringTokenizer st = new StringTokenizer(value.toString());
while(st.hasMoreTokens()){
word.set(st.nextToken());
context.write(word, new IntWritable(1));
}
}
}
public static class Reduce1 extends Reducer<Text, IntWritable, Text, IntWritable>{
IntWritable result = new IntWritable();
@Override
protected void reduce(Text key, Iterable<IntWritable> values,Context context)
throws IOException, InterruptedException {
int sum = 0;
for(IntWritable val : values){
sum += val.get();
}
result.set(sum);
context.write(key, result);
}
}
// 第二个Job
public static class Map2 extends Mapper<Object, Text, Text, IntWritable>{
Text word = new Text();
@Override
protected void map(Object key, Text value,Context context)
throws IOException, InterruptedException {
StringTokenizer st = new StringTokenizer(value.toString());
while(st.hasMoreTokens()){
word.set(st.nextToken());
context.write(word, new IntWritable(1));
}
}
}
public static class Reduce2 extends Reducer<Text, IntWritable, Text, IntWritable>{
IntWritable result = new IntWritable();
@Override
protected void reduce(Text key, Iterable<IntWritable> values,Context context)
throws IOException, InterruptedException {
int sum = 0;
for(IntWritable val : values){
sum += val.get();
}
result.set(sum);
context.write(key, result);
}
}
// 第三个Job
public static class Map3 extends Mapper<Object, Text, Text, IntWritable>{
Text word = new Text();
@Override
protected void map(Object key, Text value,Context context)
throws IOException, InterruptedException {
StringTokenizer st = new StringTokenizer(value.toString());
while(st.hasMoreTokens()){
word.set(st.nextToken());
context.write(word, new IntWritable(1));
}
}
}
public static class Reduce3 extends Reducer<Text, IntWritable, Text, IntWritable>{
IntWritable result = new IntWritable();
@Override
protected void reduce(Text key, Iterable<IntWritable> values,Context context)
throws IOException, InterruptedException {
int sum = 0;
for(IntWritable val : values){
sum += val.get();
}
result.set(sum);
context.write(key, result);
}
}
public static void main(String[] args) throws IOException{
String inPath1 = "hdfs://hadoop0:9000/user/root/3D/";
String outPath1 = "hdfs://hadoop0:9000/user/root/3DZout/";
String outPath2 = "hdfs://hadoop0:9000/user/root/3DZout2/";
String outPath3 = "hdfs://hadoop0:9000/user/root/3DZout3/";
String[] inOut = {inPath1, outPath1};
Configuration conf = new Configuration();
String[] otherArgs = new GenericOptionsParser(conf, inOut).getRemainingArgs();
if (otherArgs.length < 2) {
System.err.println("Usage: wordcount <in> [<in>...] <out>");
System.exit(2);
}
// 判断输出路径是否存在,如存在先删除
FileSystem hdfs = FileSystem.get(conf);
Path findFile = new Path(outPath1);
boolean isExists = hdfs.exists(findFile);
if(isExists){
hdfs.delete(findFile, true);
}
if(hdfs.exists(new Path(outPath2))){
hdfs.delete(new Path(outPath2), true);
}
if(hdfs.exists(new Path(outPath3))){
hdfs.delete(new Path(outPath3), true);
}
// job1配置
Job job1 = Job.getInstance(conf);
job1.setJarByClass(Mode.class);
job1.setMapperClass(Map1.class);
job1.setReducerClass(Reduce1.class);
job1.setMapOutputKeyClass(Text.class);
job1.setMapOutputValueClass(IntWritable.class);
job1.setOutputKeyClass(Text.class);
job1.setOutputValueClass(IntWritable.class);
FileInputFormat.addInputPath(job1, new Path(inPath1));
FileOutputFormat.setOutputPath(job1, new Path(outPath1));
// 将job1加入控制容器
ControlledJob ctrljob1 = new ControlledJob(conf);
ctrljob1.setJob(job1);
// job2配置
Job job2 = Job.getInstance(conf);
job2.setJarByClass(Mode.class);
job2.setMapperClass(Map2.class);
job2.setReducerClass(Reduce2.class);
job2.setMapOutputKeyClass(Text.class);
job2.setMapOutputValueClass(IntWritable.class);
job2.setOutputKeyClass(Text.class);
job2.setOutputValueClass(IntWritable.class);
FileInputFormat.addInputPath(job2, new Path(inPath1));
FileOutputFormat.setOutputPath(job2, new Path(outPath2));
// 将job2加入控制容器
ControlledJob ctrljob2 = new ControlledJob(conf);
ctrljob2.setJob(job2);
// job3配置
Job job3 = Job.getInstance(conf);
job3.setJarByClass(Mode.class);
job3.setMapperClass(Map3.class);
job3.setReducerClass(Reduce3.class);
job3.setMapOutputKeyClass(Text.class);
job3.setMapOutputValueClass(IntWritable.class);
job3.setOutputKeyClass(Text.class);
job3.setOutputValueClass(IntWritable.class);
FileInputFormat.addInputPath(job3, new Path(outPath2));
FileOutputFormat.setOutputPath(job3, new Path(outPath3));
ControlledJob ctrljob3 = new ControlledJob(conf);
// 设置job3依赖job1和job2
ctrljob3.addDependingJob(ctrljob1);
ctrljob3.addDependingJob(ctrljob2);
ctrljob3.setJob(job3);
// 主控制器
JobControl jobCtrl = new JobControl("myctrl");
jobCtrl.addJob(ctrljob1);
jobCtrl.addJob(ctrljob2);
jobCtrl.addJob(ctrljob3);
// 在启动线程,记住一定要有这个
Thread t = new Thread(jobCtrl);
t.start();
while(true){
// 如果作业全部完成,就打印成功作业的信息
if(jobCtrl.allFinished()){
System.out.println(jobCtrl.getSuccessfulJobList());
jobCtrl.stop();
break;
}
}
}
}
3、链式组合式
所谓连式MapReduce就是用多个Mapper处理任务,最后用一个Reducer输出结果,注意和迭代式和组合式MapReduce的不同之处
一个MapReduce作业可能会有一些前处理和后处理步骤,将这些前后处理步骤以单独的MapReduce任务实现也可以达到目的,但由于
增加了多个MapReduce作业,将增加整个作业的处理周期,而且还会增加很多I/O操作,因此处理效率不高。
Hadoop为此提供了专门的链式Mapper(ChainMapper)和链式Reducer(ChainReducer)来完成这种处理。
ChainMapper允许在一个单一Map任务中添加和使用多个Map子任务;而ChainReducer则允许在一个单一Reduce任务执行了Reduce处理
后,继续使用多个Map子任务完成一些后续处理。
package com.hadoop.mapreduce;
import java.io.IOException;
import java.util.StringTokenizer;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.FileSystem;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.IntWritable;
import org.apache.hadoop.io.LongWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Job;
import org.apache.hadoop.mapreduce.Mapper;
import org.apache.hadoop.mapreduce.Reducer;
import org.apache.hadoop.mapreduce.lib.chain.ChainMapper;
import org.apache.hadoop.mapreduce.lib.chain.ChainReducer;
import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;
import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;
import org.apache.hadoop.util.GenericOptionsParser;
public class Chain {
// 第一个Job
public static class Map1 extends Mapper<LongWritable, Text, Text, IntWritable>{
Text word = new Text();
@Override
protected void map(LongWritable key, Text value,Context context)
throws IOException, InterruptedException {
StringTokenizer st = new StringTokenizer(value.toString());
while(st.hasMoreTokens()){
word.set(st.nextToken());
context.write(word, new IntWritable(1));
}
}
}
public static class Reduce1 extends Reducer<Text, IntWritable, Text, IntWritable>{
IntWritable result = new IntWritable();
@Override
protected void reduce(Text key, Iterable<IntWritable> values,Context context)
throws IOException, InterruptedException {
int sum = 0;
for(IntWritable val : values){
sum += val.get();
}
result.set(sum);
context.write(key, result);
}
}
// 第二个Job
public static class Map2 extends Mapper<Text, IntWritable, Text, IntWritable>{
Text word = new Text();
@Override
protected void map(Text key, IntWritable value,Context context)
throws IOException, InterruptedException {
StringTokenizer st = new StringTokenizer(value.toString());
while(st.hasMoreTokens()){
word.set(st.nextToken());
context.write(word, new IntWritable(1));
}
}
}
public static class Reduce2 extends Reducer<Text, IntWritable, Text, IntWritable>{
IntWritable result = new IntWritable();
@Override
protected void reduce(Text key, Iterable<IntWritable> values,Context context)
throws IOException, InterruptedException {
int sum = 0;
for(IntWritable val : values){
sum += val.get();
}
result.set(sum);
context.write(key, result);
}
}
// 第三个Job
public static class Map3 extends Mapper<Text, IntWritable, Text, IntWritable>{
Text word = new Text();
@Override
protected void map(Text key, IntWritable value,Context context)
throws IOException, InterruptedException {
StringTokenizer st = new StringTokenizer(value.toString());
while(st.hasMoreTokens()){
word.set(st.nextToken());
context.write(word, new IntWritable(1));
}
}
}
public static class Reduce3 extends Reducer<Text, IntWritable, Text, IntWritable>{
IntWritable result = new IntWritable();
@Override
protected void reduce(Text key, Iterable<IntWritable> values,Context context)
throws IOException, InterruptedException {
int sum = 0;
for(IntWritable val : values){
sum += val.get();
}
result.set(sum);
context.write(key, result);
}
}
public static void main(String[] args) throws IOException, ClassNotFoundException, InterruptedException{
String inPath1 = "hdfs://hadoop0:9000/user/root/input/";
String outPath1 = "hdfs://hadoop0:9000/user/root/3DZout/";
String outPath2 = "hdfs://hadoop0:9000/user/root/3DZout2/";
String outPath3 = "hdfs://hadoop0:9000/user/root/3DZout3/";
String[] inOut = {inPath1, outPath1};
Configuration conf = new Configuration();
String[] otherArgs = new GenericOptionsParser(conf, inOut).getRemainingArgs();
if (otherArgs.length < 2) {
System.err.println("Usage: wordcount <in> [<in>...] <out>");
System.exit(2);
}
// 判断输出路径是否存在,如存在先删除
FileSystem hdfs = FileSystem.get(conf);
Path findFile = new Path(outPath1);
boolean isExists = hdfs.exists(findFile);
if(isExists){
hdfs.delete(findFile, true);
}
if(hdfs.exists(new Path(outPath2))){
hdfs.delete(new Path(outPath2), true);
}
if(hdfs.exists(new Path(outPath3))){
hdfs.delete(new Path(outPath3), true);
}
// job1配置
Job job1 = Job.getInstance(conf);
job1.setJarByClass(Chain.class);
job1.setJobName("ChainJob");
FileInputFormat.addInputPath(job1, new Path(inPath1));
FileOutputFormat.setOutputPath(job1, new Path(outPath1));
// 连式编程要注意的是,可以有多个个Mapper,且后面Mapper的输入是是上一个Mapper的输出,最后一个Mapper的输出是Reducer的输入,
// 但全局只有一个Reducer
ChainMapper.addMapper(job1, Map1.class, LongWritable.class, Text.class, Text.class, IntWritable.class, conf);
ChainMapper.addMapper(job1, Map2.class, Text.class, IntWritable.class, Text.class, IntWritable.class, conf);
ChainMapper.addMapper(job1, Map3.class, Text.class, IntWritable.class, Text.class, IntWritable.class, conf);
// 执行顺序 map1 --> map2 --> map3 --> reduce1
ChainReducer.setReducer(job1, Reduce1.class, Text.class, IntWritable.class, Text.class, IntWritable.class, conf);
job1.waitForCompletion(true);
}
}