hbase0.98 coprocessor Endpoint 之 HelloWorld

时间 2019-11-06

标签 hbase0.98 hbase coprocessor endpoint helloworld 栏目 Hadoop 繁體版

原文原文链接

###介绍： 源博地址: http://www.cockybook.com/?p=35html

HBase做为列族数据库最常常被人诟病的特性包括：没法轻易创建“二级索引”，难以执行求和、计数、排序等操做。好比，在旧版本的(<0.92)Hbase中，统计数据表的总行数，须要使用Counter方法，执行一次MapReduce Job才能获得。虽然HBase在数据存储层中集成了MapReduce，可以有效用于数据表的分布式计算。然而在不少状况下，作一些简单的相加或者聚合计算的时候，若是直接将计算过程放置在server端，可以减小通信开销，从而得到很好的性能提高。因而，HBase在0.92以后引入了协处理器(coprocessors)，实现一些激动人心的新特性：可以轻易创建二次索引、复杂过滤器(谓词下推)以及访问控制等。 HBase协处理器的灵感来自于Jeff Dean 09年的演讲( P66-67)。java

####hbase coprocessor 大类分为两种coprocessor分别是：web

RegionObserver ：它是一种相似于传统数据库的触发器，提供了钩子函数：Get、Put、Delete、Scan等。

Endpoint：是一个远程rpc调用，相似于webservice形式调用，但他不适用xml，而是使用的序列化框架是protobuf（序列化后数据更小），本文将介绍此种Coprocessor.

Endpoint 容许您定义本身的动态RPC协议,用于客户端与region servers通信。Coprocessor 与region server在相同的进程空间中，所以您能够在region端定义本身的方法（endpoint），将计算放到region端，减小网络开销，经常使用于提高hbase的功能，如：count，sum等。shell

###个人环境数据库

hadoop ： 2.2
hbase-hadoop2 ：0.98+
JDK：1.6 ##这里必需要1.6 要否则会出现不能加载jar包的现象。
操做系统：CentOS 6.4

###编写代码apache

首先你须要利用protobuf（网上本身搜google维护的目前发展到2.5版本）工具成一个HelloWorld 序列化对象。api

####HelloWorld.proto网络

option java_package = "com.gzhdi.coprocessor.generated";
option java_outer_classname = "ServerHelloworld";
option java_generic_services = true;
option java_generate_equals_and_hash = true;
option optimize_for = SPEED;

message HelloRequest {
  required bytes askWord = 10;
}

message HelloResponse {
  required bytes retWord = 10;
}

message AskRequest {
  required bytes ask = 100;
}

message AnsResponse {
  required bytes ans = 100;
}

service HelloWorld {
  rpc sendHello(HelloRequest)
    returns (HelloResponse);

  rpc question(AskRequest)
    returns (AnsResponse);
}

使用命令生成代码，并拷贝到你的工程里边去,个人文件在工程下面放着呢，直接生成到工程里边。这段代码就会生成一个HelloWorld.java文件.框架
```
protoc.exe  --java_out=../src HelloWorld.proto
```

编写主要代码分布式

####server端代码

package com.gzhdi.copocessor;

import java.io.IOException;

import org.apache.hadoop.hbase.Coprocessor;
import org.apache.hadoop.hbase.CoprocessorEnvironment;
import org.apache.hadoop.hbase.coprocessor.CoprocessorException;
import org.apache.hadoop.hbase.coprocessor.CoprocessorService;
import org.apache.hadoop.hbase.coprocessor.RegionCoprocessorEnvironment;

import com.google.protobuf.ByteString;
import com.google.protobuf.RpcCallback;
import com.google.protobuf.RpcController;
import com.google.protobuf.Service;
import com.gzhdi.coprocessor.generated.ServerHelloworld;
import com.gzhdi.coprocessor.generated.ServerHelloworld.AnsResponse;
import com.gzhdi.coprocessor.generated.ServerHelloworld.AskRequest;
import com.gzhdi.coprocessor.generated.ServerHelloworld.HelloRequest;
import com.gzhdi.coprocessor.generated.ServerHelloworld.HelloResponse;

public class HelloWorldEndPoint  extends ServerHelloworld.HelloWorld implements Coprocessor,CoprocessorService{

	private RegionCoprocessorEnvironment env; 

	@Override
	public void sendHello(RpcController controller, HelloRequest request,
			RpcCallback<HelloResponse> done) {
		System.out.println("request HelloRequest:"+request.getAskWord());
		HelloResponse resp=HelloResponse.newBuilder().setRetWord(ByteString.copyFromUtf8("hello world!!!")).build();

		done.run(resp);
	}

	@Override
	public void question(RpcController controller, AskRequest request,
			RpcCallback<AnsResponse> done) {
		System.out.println("request question:"+request.getAsk());
		AnsResponse resp=AnsResponse.newBuilder().setAns(ByteString.copyFromUtf8("helloworld,"+request.getAsk().toStringUtf8())).build();
		done.run(resp);
	}

	@Override
	public Service getService() {
		return this;
	}

	@Override
	public void start(CoprocessorEnvironment env) throws IOException {
		if (env instanceof RegionCoprocessorEnvironment) {  
		      this.env = (RegionCoprocessorEnvironment)env;  
		    } else {  
		      throw new CoprocessorException("Must be loaded on a table region!");  
		    }  
	}

	@Override
	public void stop(CoprocessorEnvironment env) throws IOException {

	}
}

####client 端代码

package com.gzhdi.copocessor;

import java.io.IOException;
import java.util.Map;

import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.hbase.HBaseConfiguration;
import org.apache.hadoop.hbase.client.HTable;
import org.apache.hadoop.hbase.client.coprocessor.Batch;
import org.apache.hadoop.hbase.ipc.BlockingRpcCallback;
import org.apache.hadoop.hbase.ipc.ServerRpcController;

import com.google.protobuf.ByteString;
import com.google.protobuf.ServiceException;
import com.gzhdi.coprocessor.generated.ServerHelloworld.AnsResponse;
import com.gzhdi.coprocessor.generated.ServerHelloworld.AskRequest;
import com.gzhdi.coprocessor.generated.ServerHelloworld.HelloWorld;

public class HelloWorldClient {

	public static void main(String[] args) throws ServiceException, Throwable {
		myclient();
	}
//若是你没有写好本身的例子能够跑跑hbase自带的小例子
//	private static void example1() throws IOException, ServiceException,
//			Throwable {
//		System.out.println("begin.....");  
//        long begin_time=System.currentTimeMillis();  
//       Configuration config=HBaseConfiguration.create();  
////     String master_ip="192.168.150.128";  
//       String master_ip="10.10.113.211";  
//       String zk_ip="10.10.113.211";  
//       String table_name="t1";  
//       config.set("hbase.zookeeper.property.clientPort", "2181");   
//       config.set("hbase.zookeeper.quorum", zk_ip);   
//       config.set("hbase.master", master_ip+":600000");  
//       
//       HTable table = new HTable(config, table_name);
//       final ExampleProtos.CountRequest request = ExampleProtos.CountRequest.getDefaultInstance();
//       Map results = table.coprocessorService(
//           ExampleProtos.RowCountService.class, // the protocol interface we're invoking
//           null, null,                          // start and end row keys
//           
//           new Batch.Call() {
//        	   
//               public Long call(Object counter) throws IOException {
//                 BlockingRpcCallback<ExampleProtos.CountResponse> rpcCallback =
//                     new BlockingRpcCallback();
//                 ((ExampleProtos.RowCountService)counter).getRowCount(null, request, rpcCallback);
//                 ExampleProtos.CountResponse response = rpcCallback.get();
//                 System.out.println("count :::::"+response.getCount());
//                 return response.hasCount() ? response.getCount() : 0;
//               }
//			
//           });
//	}

	public static void myclient(){
		// TODO Auto-generated method stub
				System.out.println("begin.....");  
		        long begin_time=System.currentTimeMillis();  
		       Configuration config=HBaseConfiguration.create();  
//		     String master_ip="192.168.150.128";  
		       String master_ip="10.10.113.211";  
		       String zk_ip="10.10.113.211";  
		       String table_name="t1";  
		       config.set("hbase.zookeeper.property.clientPort", "2181");   
		       config.set("hbase.zookeeper.quorum", zk_ip);   
		       config.set("hbase.master", master_ip+":600000");  

		       final AskRequest req=AskRequest.newBuilder().setAsk(ByteString.copyFromUtf8("hello")).build();
		       AnsResponse resp=null;
		       try {
				HTable table=new HTable(config,table_name);
				Map<byte[], ByteString> re=table.coprocessorService(HelloWorld.class, null, null, new Batch.Call<HelloWorld, ByteString>() {

					@Override
					public ByteString call(HelloWorld instance) throws IOException {
						ServerRpcController controller = new ServerRpcController();
						BlockingRpcCallback<AnsResponse> rpccall=new BlockingRpcCallback<AnsResponse>();
						instance.question(controller, req, rpccall);
						AnsResponse resp=rpccall.get();


						//result
						System.out.println("resp:"+ resp.getAns().toStringUtf8());

						return resp.getAns();
					}

				});
			} catch (IOException e) {
				e.printStackTrace();
			} catch (ServiceException e) {
				e.printStackTrace();
			} catch (Throwable e) {
				e.printStackTrace();
			}  
	}

}

利用jdk 1.6打包（切记jdk1.6，由于hbase用1.6打包的）导出hellworld.jar 包名随便起。

###部署

将包helloworld.jar 放在 %HBASE_HOME/lib/ 下就能够了。
从新启动hbase

验证

[root@hdp22 ~ Desktop]# hbase shell
hbase(main):001:0> import com.gzhdi.copocessor.HelloWorldEndPoint
=> Java::ComGzhdiCopocessor::HelloWorldEndPoint    //若是打印出这句话就说明包已经加载完毕

向指定表添加endpoint

hbase(main):002:0> create 't1','f1'
0 row(s) in 6.5290 seconds
=> Hbase::Table - t1   //建立表t1

hbase(main):003:0> alter 't1','coprocessor'=>'|com.gzhdi.copocessor.HelloWorldEndPoint|1001|'
Updating all regions with the new schema...
0/1 regions updated.
1/1 regions updated.
Done.
0 row(s) in 2.5960 seconds

hbase(main):005:0> describe 't1'
DESCRIPTION                                                                                                              ENABLED                                                          
 't1', {TABLE_ATTRIBUTES => {coprocessor$1 => '|com.gzhdi.copocessor.HelloWorldEndPoint|1001|'}, {NAME => 'f1', DATA_BLO true                                                             
 CK_ENCODING => 'NONE', BLOOMFILTER => 'ROW', REPLICATION_SCOPE => '0', VERSIONS => '1', COMPRESSION => 'NONE', MIN_VERS                                                                  
 IONS => '0', TTL => '2147483647', KEEP_DELETED_CELLS => 'false', BLOCKSIZE => '65536', IN_MEMORY => 'false', BLOCKCACHE                                                                  
  => 'true'}                                                                                                                                                                              
1 row(s) in 0.0940 seconds

//OK 成功了

###调用如今就可使用你的客户端代码调用该服务了，须要制定zookeeper地址和表名（由于服务是针对表的）。

###参考资料 http://blog.csdn.net/yangzongzhao/article/details/24306775 https://blogs.apache.org/hbase/entry/coprocessor_introduction#comment-1337006507000 http://blog.csdn.net/hljlzc2007/article/details/12652243 http://hbase.apache.org/apidocs/org/apache/hadoop/hbase/client/coprocessor/package-summary.html http://hbase.apache.org/apidocs/org/apache/hadoop/hbase/coprocessor/package-summary.html