hbase0.98 coprocessor Endpoint 之 HelloWorld

介绍:

源博地址: http://www.cockybook.com/?p=35

HBase作为列族数据库最经常被人诟病的特性包括:无法轻易建立“二级索引”,难以执行求和、计数、排序等操作。比如,在旧版本的(<0.92)Hbase中,统计数据表的总行数,需要使用Counter方法,执行一次MapReduce Job才能得到。虽然HBase在数据存储层中集成了MapReduce,能够有效用于数据表的分布式计算。然而在很多情况下,做一些简单的相加或者聚合计算的时候,如果直接将计算过程放置在server端,能够减少通讯开销,从而获得很好的性能提升。于是,HBase在0.92之后引入了协处理器(coprocessors),实现一些激动人心的新特性:能够轻易建立二次索引、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)。

hbase coprocessor 大类分为两种coprocessor分别是:

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

  2. Endpoint:是一个远程rpc调用,类似于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 允许您定义自己的动态RPC协议,用于客户端与region servers通讯。Coprocessor 与region server在相同的进程空间中,因此您可以在region端定义自己的方法(endpoint),将计算放到region端,减少网络开销,常用于提升hbase的功能,如:count,sum等。

我的环境

  • hadoop : 2.2
  • hbase-hadoop2 :0.98+
  • JDK:1.6 ##这里必须要1.6 要不然会出现不能加载jar包的现象。
  • 操作系统:CentOS 6.4

编写代码

  1. 首先你需要利用protobuf(网上自己搜google维护的目前发展到2.5版本) 工具成一个HelloWorld 序列化对象。

    HelloWorld.proto

    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);
    }
    
  2. 使用命令生成代码,并拷贝到你的工程里边去,我的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.

    protoc.exe  --java_out=../src HelloWorld.proto
    
  3. 编写主要代码

    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();
        }
    //如果你没有写好自己的例子可以跑跑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();
                }  
        }
    
    }
    
  4. 利用jdk 1.6打包(切记jdk1.6,因为hbase用1.6打包的) 导出hellworld.jar 包名随便起。

部署

  1. 将包helloworld.jar 放在 %HBASE_HOME/lib/ 下就可以了。

  2. 重新启动hbase

  3. 验证

     [root@hdp22 ~ Desktop]# hbase shell
    hbase(main):001:0> import com.gzhdi.copocessor.HelloWorldEndPoint
    => Java::ComGzhdiCopocessor::HelloWorldEndPoint    //如果打印出这句话就说明包已经加载完毕
    
  4. 向指定表添加endpoint

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

现在就可以使用你的客户端代码调用该服务了,需要制定zookeeper地址和表名(因为服务是针对表的)。

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

你可能感兴趣的:(hbase,endpoint,coprocessor)