大家好，我是Iggi。
今天我给大家分享的是MapReduce2-3.1.1版本的Join实验。
关于MapReduce的一段文字简介请自行查阅我的实验示例：MapReduce2-3.1.1 实验示例 单词计数（一）
好，下面进入正题。介绍Java操作MapReduce2组件完成两个结果集的Join的操作。
首先，使用IDE建立Maven工程，建立工程时没有特殊说明，按照向导提示点击完成即可。重要的是在pom.xml文件中添加依赖包，内容如下图：
image.png
待系统下载好依赖的jar包后便可以编写程序了。
展示实验代码：
package linose.mapreduce.join;
import java.io.IOException;
import java.io.OutputStreamWriter;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.FSDataInputStream;
import org.apache.hadoop.fs.FSDataOutputStream;
import org.apache.hadoop.fs.FileSystem;
import org.apache.hadoop.fs.LocatedFileStatus;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.fs.RemoteIterator;
import org.apache.hadoop.io.IOUtils;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Job;
import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;
import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;
//import org.apache.log4j.BasicConfigurator;
public class AppJoin {
public static void main(String[] args) throws IOException, ClassNotFoundException, InterruptedException {
/**
* 设定MapReduce示例拥有HDFS的操作权限
*/
System.setProperty("HADOOP_USER_NAME", "hdfs");
/**
* 为了清楚的看到输出结果，暂将集群调试信息缺省。
* 如果想查阅集群调试信息，取消注释即可。
*/
//BasicConfigurator.configure();
/**
* MapReude实验准备阶段：
* 定义HDFS文件路径
*/
String defaultFS = "hdfs://master2.linose.cloud.beijing.com:8020";
String inputPath1 = defaultFS + "/index.dirs/data1.txt";
String inputPath2 = defaultFS + "/index.dirs/data2.txt";
String outputPath = defaultFS + "/index.dirs/joinresult.txt";
/**
* 定义输入路径，输出路径
*/
Path inputHdfsPath1 = new Path(inputPath1);
Path inputHdfsPath2 = new Path(inputPath2);
Path outputHdfsPath = new Path(outputPath);
/**
* 生产配置，并获取HDFS对象
*/
Configuration conf = new Configuration();
conf.set("fs.defaultFS", defaultFS);
FileSystem system = FileSystem.get(conf);
/**
* 如果实验数据文件不存在则创建数据文件
*/
system.delete(inputHdfsPath1, false);
if (!system.exists(inputHdfsPath1)) {
FSDataOutputStream outputStream = system.create(inputHdfsPath1);
OutputStreamWriter file = new OutputStreamWriter(outputStream);
file.write("2019-05-13 13:00\t001\tIggi\tbeijing\t010\n");
file.write("2019-05-13 14:00\t002\tchap\tbeijing\t010\n");
file.write("2019-05-13 15:00\t003\ttxingsg\tbeijing\t010\n");
file.write("2019-05-13 16:00\t004\tthedd\tbeijing\t010\n");
file.write("2019-05-13 17:00\t005\ttlix\tbeijing\t010\n");
file.close();
outputStream.close();
}
system.delete(inputHdfsPath2, false);
if (!system.exists(inputHdfsPath2)) {
FSDataOutputStream outputStream = system.create(inputHdfsPath2);
OutputStreamWriter file = new OutputStreamWriter(outputStream);
file.write("002\tjava\n");
file.write("003\tjava\n");
file.write("004\tproduct\n");
file.write("005\tmanager\n");
file.close();
outputStream.close();
}
/**
* 如果实验结果目录存在，遍历文件内容全部删除
*/
if (system.exists(outputHdfsPath)) {
RemoteIterator fsIterator = system.listFiles(outputHdfsPath, true);
LocatedFileStatus fileStatus;
while (fsIterator.hasNext()) {
fileStatus = fsIterator.next();
system.delete(fileStatus.getPath(), false);
}
system.delete(outputHdfsPath, false);
}
/**
* 创建MapReduce任务并设定Job名称
*/
Job job = Job.getInstance(conf, "Join Data Set");
job.setJarByClass(JoinMapper.class);
/**
* 设置输入文件、输出文件
*/
FileInputFormat.addInputPath(job, inputHdfsPath1);
FileInputFormat.addInputPath(job, inputHdfsPath2);
FileOutputFormat.setOutputPath(job, outputHdfsPath);
/**
* 指定Reduce类输出类型Key类型与Value类型
*/
job.setOutputKeyClass(Text.class);
job.setOutputValueClass(Text.class);
/**
* 指定Map类输出类型Key类型与Value类型
*/
job.setMapOutputKeyClass(TextPair.class);
job.setMapOutputValueClass(Text.class);
/**
* 指定自定义Map类，Reduce类，Partitioner类、Comparator类
*/
job.setMapperClass(JoinMapper.mrJoinMapper.class);
job.setReducerClass(JoinMapper.mrJoinReduce.class);
job.setPartitionerClass(JoinMapper.mrJoinPartitioner.class);
job.setGroupingComparatorClass(JoinMapper.mrJoinComparator.class);
/**
* 提交作业
*/
job.waitForCompletion(true);
/**
* 然后轮询进度，直到作业完成。
*/
float progress = 0.0f;
do {
progress = job.setupProgress();
System.out.println("Word Count Ver2: 的当前进度：" + progress * 100);
Thread.sleep(1000);
} while (progress != 1.0f && !job.isComplete());
/**
* 如果成功，查看输出文件内容
*/
if (job.isSuccessful()) {
RemoteIterator fsIterator = system.listFiles(outputHdfsPath, true);
LocatedFileStatus fileStatus;
while (fsIterator.hasNext()) {
fileStatus = fsIterator.next();
FSDataInputStream outputStream = system.open(fileStatus.getPath());
IOUtils.copyBytes(outputStream, System.out, conf, false);
outputStream.close();
System.out.println("--------------------------------------------");
}
}
}
}
展示MapReduce2-3.1.1组件编写TextPair测试类：
package linose.mapreduce.join;
import java.io.DataInput;
import java.io.DataOutput;
import java.io.IOException;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.io.WritableComparable;
import linose.mapreduce.secondarysort.IntPair;
public class TextPair implements WritableComparable{
/**
* 定义：指定键、匹配键
*/
private Text pairKey;
private Text specificKey;
public TextPair() {
set(new Text(), new Text());
}
public TextPair(Text pairKey, Text specificKey) {
set(pairKey, specificKey);
}
public TextPair(String pairKey, String specificKey) {
set(new Text(pairKey), new Text(specificKey));
}
public void set(Text pairKey, Text specificKey) {
this.pairKey = pairKey;
this.specificKey = specificKey;
}
public Text getPairKey() {
return pairKey;
}
public Text getSpecificKey() {
return specificKey;
}
public void readFields(DataInput input) throws IOException {
pairKey.readFields(input);
specificKey.readFields(input);
}
public void write(DataOutput output) throws IOException {
pairKey.write(output);
specificKey.write(output);
}
public int compareTo(TextPair o) {
int compare = pairKey.compareTo(o.pairKey);
if (0 != compare) {
return compare;
}
return specificKey.compareTo(o.specificKey);
}
public int hashCode() {
return pairKey.hashCode()*163 + specificKey.hashCode();
}
public boolean equals(Object o) {
if (o instanceof IntPair) {
TextPair pair = (TextPair)o;
return pairKey.equals(pair.pairKey) && specificKey.equals(pair.specificKey);
}
return false;
}
public String toString() {
return pairKey + "\t" + specificKey;
}
}
展示MapReduce2-3.1.1组件编写Join测试类：
package linose.mapreduce.join;
import java.io.IOException;
import org.apache.hadoop.io.LongWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.io.WritableComparable;
import org.apache.hadoop.io.WritableComparator;
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.FileSplit;
public class JoinMapper {
public static class mrJoinMapper extends Mapper {
protected void map(LongWritable key, Text value, Context context) throws IOException, InterruptedException {
FileSplit file = (FileSplit) context.getInputSplit();
String path = file.getPath().toString();
if (path.contains("data1.txt")) {
String values[] = value.toString().split("\t");
if (values.length < 3) {
return;
}
String SpecificKey = values[0];
String pairKey = values[1];
StringBuffer buffer = new StringBuffer();
buffer.append(SpecificKey);
buffer.append("\t");
for (int i = 2; i < values.length; ++i) {
buffer.append(values[i]);
buffer.append("\t");
}
TextPair outputKey = new TextPair(pairKey, "1");
Text outputValue = new Text(buffer.toString());
context.write(outputKey, outputValue);
}
if (path.contains("data2.txt")) {
String values[] = value.toString().split("\t");
if (values.length < 2) {
return;
}
String pairKey = values[0];
StringBuffer buffer = new StringBuffer();
for (int i = 1; i < values.length; ++i) {
buffer.append(values[i]);
buffer.append("\t");
}
TextPair outputKey = new TextPair(pairKey, "0");
Text outputValue = new Text(buffer.toString());
context.write(outputKey, outputValue);
}
}
}
public static class mrJoinReduce extends Reducer {
protected void reduce(TextPair key, Iterable values, Context context) throws IOException, InterruptedException {
Text pairKey = key.getPairKey();
String val = values.iterator().next().toString();
while (values.iterator().hasNext()) {
Text pairValue = new Text(values.iterator().next().toString() + "\t" + val);
context.write(pairKey, pairValue);
}
}
}
public static class mrJoinPartitioner extends Partitioner {
@Override
public int getPartition(TextPair key, Text value, int partition) {
return Math.abs(key.getPairKey().hashCode() * 127) % partition;
}
}
public static class mrJoinComparator extends WritableComparator {
public mrJoinComparator() {
super(TextPair.class, true);
}
public int compare(@SuppressWarnings("rawtypes") WritableComparable data1, @SuppressWarnings("rawtypes") WritableComparable data2) {
TextPair d1 = (TextPair)data1;
TextPair d2 = (TextPair)data2;
return d1.getPairKey().compareTo(d2.getPairKey());
}
}
}
下图为测试结果：
image.png
至此，MapReduce2-3.1.1 Join实验示例演示完毕。