Hbase 与mapreduce结合

Hbase和mapreduce结合

为什么需要用mapreduce去访问hbase的数据？

——加快分析速度和扩展分析能力

Mapreduce访问hbase数据作分析一定是在离线分析的场景下应用

案例1、HBase表数据的转移

在Hadoop阶段，我们编写的MR任务分别进程了Mapper和Reducer两个类，而在HBase中我们需要继承的是TableMapper和TableReducer两个类。

目标：将fruit表中的一部分数据，通过MR迁入到fruit_mr表中

Step1、构建ReadFruitMapper类，用于读取fruit表中的数据

---

import java.io.IOException;

import org.apache.hadoop.hbase.Cell;

import org.apache.hadoop.hbase.CellUtil;

import org.apache.hadoop.hbase.client.Put;

import org.apache.hadoop.hbase.client.Result;

import org.apache.hadoop.hbase.io.ImmutableBytesWritable;

import org.apache.hadoop.hbase.mapreduce.TableMapper;

import org.apache.hadoop.hbase.util.Bytes;

public class ReadFruitMapper extends TableMapper<ImmutableBytesWritable, Put> {

@Override

protected void map(ImmutableBytesWritable key, Result value, Context context)

throws IOException, InterruptedException {

//将fruit的name和color提取出来，相当于将每一行数据读取出来放入到Put对象中。

Put put = new Put(key.get());

//遍历添加column行

for(Cell cell: value.rawCells()){

//添加/克隆列族:info

if("info".equals(Bytes.toString(CellUtil.cloneFamily(cell)))){

//添加/克隆列：name

if("name".equals(Bytes.toString(CellUtil.cloneQualifier(cell)))){

//将该列cell加入到put对象中

put.add(cell);

//添加/克隆列:color

}else if("color".equals(Bytes.toString(CellUtil.cloneQualifier(cell)))){

//向该列cell加入到put对象中

put.add(cell);

}

}

}

//将从fruit读取到的每行数据写入到context中作为map的输出

context.write(key, put);

}

}

---

Step2、构建WriteFruitMRReducer类，用于将读取到的fruit表中的数据写入到fruit_mr表中

---

import java.io.IOException;

import org.apache.hadoop.hbase.client.Put;

import org.apache.hadoop.hbase.io.ImmutableBytesWritable;

import org.apache.hadoop.hbase.mapreduce.TableReducer;

import org.apache.hadoop.io.NullWritable;

public class WriteFruitMRReducer extends TableReducer<ImmutableBytesWritable, Put, NullWritable> {

@Override

protected void reduce(ImmutableBytesWritable key, Iterable<Put> values, Context context)

throws IOException, InterruptedException {

//读出来的每一行数据写入到fruit_mr表中

for(Put put: values){

context.write(NullWritable.get(), put);

}

}

}

---

Step3、构建Fruit2FruitMRJob extends Configured implements Tool，用于组装运行Job任务

---

//组装Job

public int run(String[] args) throws Exception {

//得到Configuration

Configuration conf = this.getConf();

//创建Job任务

Job job = Job.getInstance(conf, this.getClass().getSimpleName());

job.setJarByClass(Fruit2FruitMRJob.class);

//配置Job

Scan scan = new Scan();

scan.setCacheBlocks(false);

scan.setCaching(500);

//设置Mapper，注意导入的是mapreduce包下的，不是mapred包下的，后者是老版本

TableMapReduceUtil.initTableMapperJob(

"fruit", //数据源的表名

scan, //scan扫描控制器

ReadFruitMapper.class,//设置Mapper类

ImmutableBytesWritable.class,//设置Mapper输出key类型

Put.class,//设置Mapper输出value值类型

job//设置给哪个JOB

);

//设置Reducer

TableMapReduceUtil.initTableReducerJob("fruit_mr", WriteFruitMRReducer.class, job);

//设置Reduce数量，最少1个

job.setNumReduceTasks(1);

boolean isSuccess = job.waitForCompletion(true);

if(!isSuccess){

throw new IOException("Job running with error");

}

return isSuccess ? 0 : 1;

}

---

Step4、主函数中调用运行该Job任务

---

public static void main( String[] args ) throws Exception{

Configuration conf = HBaseConfiguration.create();

int status = ToolRunner.run(conf, new Fruit2FruitMRJob(), args);

System.exit(status);

}

---

案例2：从Hbase中读取数据、分析，写入hdfs

/**

public abstract class TableMapper<KEYOUT, VALUEOUT>

extends Mapper<ImmutableBytesWritable, Result, KEYOUT, VALUEOUT> {

}

* @author duanhaitao@gec.cn

*

*/

public class HbaseReader {

public static String flow_fields_import = "flow_fields_import";

static class HdfsSinkMapper extends TableMapper<Text, NullWritable>{

@Override

protected void map(ImmutableBytesWritable key, Result value, Context context) throws IOException, InterruptedException {

byte[] bytes = key.copyBytes();

String phone = new String(bytes);

byte[] urlbytes = value.getValue("f1".getBytes(), "url".getBytes());

String url = new String(urlbytes);

context.write(new Text(phone + "\t" + url), NullWritable.get());

}

}

static class HdfsSinkReducer extends Reducer<Text, NullWritable, Text, NullWritable>{

@Override

protected void reduce(Text key, Iterable<NullWritable> values, Context context) throws IOException, InterruptedException {

context.write(key, NullWritable.get());

}

}

public static void main(String[] args) throws Exception {

Configuration conf = HBaseConfiguration.create();

conf.set("hbase.zookeeper.quorum", "spark01");

Job job = Job.getInstance(conf);

job.setJarByClass(HbaseReader.class);

//            job.setMapperClass(HdfsSinkMapper.class);

Scan scan = new Scan();

TableMapReduceUtil.initTableMapperJob(flow_fields_import, scan, HdfsSinkMapper.class, Text.class, NullWritable.class, job);

job.setReducerClass(HdfsSinkReducer.class);

FileOutputFormat.setOutputPath(job, new Path("c:/hbasetest/output"));

job.setOutputKeyClass(Text.class);

job.setOutputValueClass(NullWritable.class);

job.waitForCompletion(true);

}

}

2.3.2 从hdfs中读取数据写入Hbase

q

/**

public abstract class TableReducer<KEYIN, VALUEIN, KEYOUT>

extends Reducer<KEYIN, VALUEIN, KEYOUT, Writable> {

}

* @author duanhaitao@gec.cn

*

*/

public class HbaseSinker {

public static String flow_fields_import = "flow_fields_import";

static class HbaseSinkMrMapper extends Mapper<LongWritable, Text, FlowBean, NullWritable>{

@Override

protected void map(LongWritable key, Text value, Context context) throws IOException, InterruptedException {

String line = value.toString();

String[] fields = line.split("\t");

String phone = fields[0];

String url = fields[1];

FlowBean bean = new FlowBean(phone,url);

context.write(bean, NullWritable.get());

}

}

static class HbaseSinkMrReducer extends TableReducer<FlowBean, NullWritable, ImmutableBytesWritable>{

@Override

protected void reduce(FlowBean key, Iterable<NullWritable> values, Context context) throws IOException, InterruptedException {

Put put = new Put(key.getPhone().getBytes());

put.add("f1".getBytes(), "url".getBytes(), key.getUrl().getBytes());

context.write(new ImmutableBytesWritable(key.getPhone().getBytes()), put);

}

}

public static void main(String[] args) throws Exception {

Configuration conf = HBaseConfiguration.create();

conf.set("hbase.zookeeper.quorum", "spark01");

HBaseAdmin hBaseAdmin = new HBaseAdmin(conf);

boolean tableExists = hBaseAdmin.tableExists(flow_fields_import);

if(tableExists){

hBaseAdmin.disableTable(flow_fields_import);

hBaseAdmin.deleteTable(flow_fields_import);

}

HTableDescriptor desc = new HTableDescriptor(TableName.valueOf(flow_fields_import));

HColumnDescriptor hColumnDescriptor = new HColumnDescriptor ("f1".getBytes());

desc.addFamily(hColumnDescriptor);

hBaseAdmin.createTable(desc);

Job job = Job.getInstance(conf);

job.setJarByClass(HbaseSinker.class);

job.setMapperClass(HbaseSinkMrMapper.class);

TableMapReduceUtil.initTableReducerJob(flow_fields_import, HbaseSinkMrReducer.class, job);

FileInputFormat.setInputPaths(job, new Path("c:/hbasetest/data"));

job.setMapOutputKeyClass(FlowBean.class);

job.setMapOutputValueClass(NullWritable.class);

job.setOutputKeyClass(ImmutableBytesWritable.class);

job.setOutputValueClass(Mutation.class);

job.waitForCompletion(true);

}

}