基于开源Dubbo分布式RPC服务框架的部署整合

一、前言

Dubbo 作为SOA服务化治理方案的核心框架，用于提高业务逻辑的复用、整合、集中管理，具有极高的可靠性（HA）和伸缩性，被应用于阿里巴巴各成员站点，同时在包括JD、当当在内的众多互联网项目中有着广泛应用。dubbo 通过高性能 RPC 实现服务的输出和输入功能，框架基于 Spring Framework 进行无缝集成，使用过程中基本看不到 Dubbo API的直接调用，Dubbo服务支持RMI、Hessian、Dubbo、WebService等众多通信协议，同时提供了对服务的监控和管理平台，属于一套完整的SOA解决方案。

详细介绍请参照官方地址：  ，不再重复描述，本文主要记录了详细的开发整合步骤。 引用一张官方提供的dubbo架构图：

• Provider    暴露服务方称之为“服务提供者”。

• Consumer 调用远程服务方称之为“服务消费者”。

• Registry    服务注册与发现的中心目录服务称之为“服务注册中心”。

• Monitor     统计服务的调用次调和调用时间的日志服务称之为“服务监控中心”。

• Container  服务运行“容器”。

二、源码构建

1、dubbo基于maven构建，需要先配置maven环境，当然如果你不想构建源码，也可以直接下载对应版本的发行包。

2、dubbo构建需要依赖opensesame组件，先通过 git 获取并安装opensesame，源码地址：

https://github.com/alibaba/opensesame

执行maven安装：mvn install

3、通过下面 git 地址获取对应版本的 dubbo 源码，官方推荐使用 dubbo 2.4.9，本文基于trunk版本构建。

https://github.com/alibaba/dubbo

maven构建：mvn clean install -Dmaven.test.skip，在漫长的等待后（过程需要通过网络下载依赖资源）会有如下提示，生成的 jar 包位于各自目录的target文件夹。

[INFO] ------------------------------------------------------------------------ [INFO] Reactor Summary:
[INFO]
[INFO] dubbo-parent ....................................... SUCCESS [ 1.272 s]
[INFO] dubbo-common ....................................... SUCCESS [ 4.572 s]
[INFO] dubbo-container .................................... SUCCESS [ 0.024 s]
[INFO] dubbo-container-api ................................ SUCCESS [ 1.148 s]
[INFO] dubbo-container-spring ............................. SUCCESS [ 0.524 s]
[INFO] dubbo-container-jetty .............................. SUCCESS [ 0.420 s]
[INFO] dubbo-container-log4j .............................. SUCCESS [ 0.448 s]
[INFO] dubbo-container-logback ............................ SUCCESS [ 0.468 s]
[INFO] dubbo-remoting ..................................... SUCCESS [ 0.016 s]
[INFO] dubbo-remoting-api ................................. SUCCESS [ 1.864 s]
[INFO] dubbo-remoting-netty ............................... SUCCESS [ 1.125 s]
[INFO] dubbo-remoting-mina ................................ SUCCESS [ 0.696 s]
[INFO] dubbo-remoting-grizzly ............................. SUCCESS [ 0.680 s]
[INFO] dubbo-remoting-p2p ................................. SUCCESS [ 1.048 s]
[INFO] dubbo-remoting-http ................................ SUCCESS [ 0.616 s]
[INFO] dubbo-remoting-zookeeper ........................... SUCCESS [ 0.716 s]
[INFO] dubbo-rpc .......................................... SUCCESS [ 0.016 s]
[INFO] dubbo-rpc-api ...................................... SUCCESS [ 1.608 s]
[INFO] dubbo-rpc-default .................................. SUCCESS [ 1.396 s]
[INFO] dubbo-rpc-injvm .................................... SUCCESS [ 0.760 s]
[INFO] dubbo-rpc-rmi ...................................... SUCCESS [ 0.416 s]
[INFO] dubbo-rpc-hessian .................................. SUCCESS [ 0.584 s]
[INFO] dubbo-rpc-http ..................................... SUCCESS [ 0.580 s]
[INFO] dubbo-rpc-webservice ............................... SUCCESS [ 0.536 s]
[INFO] dubbo-cluster ...................................... SUCCESS [ 2.208 s]
[INFO] dubbo-registry ..................................... SUCCESS [ 0.020 s]
[INFO] dubbo-registry-api ................................. SUCCESS [ 1.241 s]
[INFO] dubbo-monitor ...................................... SUCCESS [ 0.016 s]
[INFO] dubbo-monitor-api .................................. SUCCESS [ 0.464 s]
[INFO] dubbo-filter ....................................... SUCCESS [ 0.016 s]
[INFO] dubbo-filter-validation ............................ SUCCESS [ 0.608 s]
[INFO] dubbo-filter-cache ................................. SUCCESS [ 0.604 s]
[INFO] dubbo-registry-default ............................. SUCCESS [ 0.540 s]
[INFO] dubbo-monitor-default .............................. SUCCESS [ 0.588 s]
[INFO] dubbo-registry-multicast ........................... SUCCESS [ 0.632 s]
[INFO] dubbo-config ....................................... SUCCESS [ 0.016 s]
[INFO] dubbo-config-api ................................... SUCCESS [ 1.500 s]
[INFO] dubbo-config-spring ................................ SUCCESS [ 1.520 s]
[INFO] dubbo-rpc-thrift ................................... SUCCESS [ 1.032 s]
[INFO] dubbo-rpc-memcached ................................ SUCCESS [ 0.464 s]
[INFO] dubbo-rpc-redis .................................... SUCCESS [ 0.500 s]
[INFO] dubbo-registry-zookeeper ........................... SUCCESS [ 0.656 s]
[INFO] dubbo-registry-redis ............................... SUCCESS [ 0.640 s]
[INFO] dubbo .............................................. SUCCESS [ 1.009 s]
[INFO] dubbo-simple ....................................... SUCCESS [ 0.016 s]
[INFO] dubbo-registry-simple .............................. SUCCESS [01:43 min]
[INFO] dubbo-monitor-simple ............................... SUCCESS [ 10.379 s]
[INFO] dubbo-admin ........................................ SUCCESS [04:03 min]
[INFO] dubbo-demo ......................................... SUCCESS [ 0.016 s]
[INFO] dubbo-demo-api ..................................... SUCCESS [ 0.308 s]
[INFO] dubbo-demo-provider ................................ SUCCESS [ 2.700 s]
[INFO] dubbo-demo-consumer ................................ SUCCESS [ 2.745 s]
[INFO] dubbo-test ......................................... SUCCESS [ 0.012 s]
[INFO] dubbo-test-benchmark ............................... SUCCESS [01:01 min]
[INFO] dubbo-test-compatibility ........................... SUCCESS [ 0.188 s]
[INFO] dubbo-test-integration ............................. SUCCESS [ 0.052 s]
[INFO] dubbo-test-examples ................................ SUCCESS [ 1.640 s]
[INFO] ------------------------------------------------------------------------ [INFO] BUILD SUCCESS
[INFO] ------------------------------------------------------------------------ [INFO] Total time: 07:42 min
[INFO] Finished at: 2015-01-20T14:36:43+08:00 [INFO] Final Memory: 282M/895M
[INFO] ------------------------------------------------------------------------

View Code

4、 构建过程中如果出现maven网络连接相关的异常信息，可以尝试将以下maven mirror 地址添加到 settings.xml 文件。

<mirror>  <id>CNid>  <mirrorOf>centralmirrorOf>  <name>OSChina Centralname>  <url>url> mirror> <mirror>  <id>kafeituid>  <mirrorOf>centralmirrorOf>  <name>Human Readable Name for this Mirror.name>  <url>url> mirror> <mirror>  <id>lvu.cnid>  <name>lvu.cnname>  <url>url>  <mirrorOf>*mirrorOf> mirror> 

三、Zookeeper部署（Registry 服务注册中心）

D ubbo 缺省配置通过 multicast 注册中心广播实现 Provider 和 Consumer 之间的简单远程过程调用（Simple RPC），不需要通过 Registry 注册中心进行注册调度 ，类似于spring rmi remoting调用，但由于不是Cluster部署，所以作为分布 式RPC框架官方建议使用 Zookeeper 作为Registry注册中心服务器（同样支持Redis）实现服务的注册、发现、路由功能。Dubbo在Zookeeper服务器端只增加了dubbo数据节点（如下图），无需其他任何配置，所以只需安装或使用现有 Zookeeper 服务器即可，关于Zookeeper的安装部署可以参照之前的博文：http://www.cnblogs.com/lengfo/p/4146797.html ，不再重复介绍。

Zookeeper同样支持集群部署，提供最终一致性数据服务，本文实验环境只使用了一台Linux服务器作为ZK Registry注册服务器，zoo.cfg配置文件如下。

# The number of milliseconds of each tick tickTime=2000 # The number of ticks that the initial  # synchronization phase can take initLimit=10 # The number of ticks that can pass between  # sending a request and getting an acknowledgement syncLimit=5 # the directory where the snapshot is stored. # do not use /tmp for storage, /tmp here is just  # example sakes. dataDir=/home/glf/zookeeperServer/data dataLogDir=/home/glf/zookeeperServer/logs # the port at which the clients will connect clientPort=2181 # the maximum number of client connections. # increase this if you need to handle more clients #maxClientCnxns=60 # # Be sure to read the maintenance section of the  # administrator guide before turning on autopurge. # # http://zookeeper.apache.org/doc/current/zookeeperAdmin.html#sc_maintenance # # The number of snapshots to retain in dataDir #autopurge.snapRetainCount=3 # Purge task interval in hours # Set to "0" to disable auto purge feature #autopurge.purgeInterval=1

四、服务提供方开发（Provider）

1、创建Java测试项目 DubboDemo ，集成Spring Framework。由于dubbo基于Spring的Schema扩展进行加载，依赖于低版本spring，所以如果开发项目中使用高版本（spring 3.0及以上）会出现xml解析冲突，官方QA中也有相关提及： ，所以本文只集成了2.5.6版本spring，更多相关版本问题也有待进一步研究，Todo。

2、集成dubbo及相关依赖包，参照下图，如果使用dubbo其他特性可能需要依赖更多包，比如jedis。

3、创建测试服务接口（IGoodsManager） 

package ibusiness; public interface IGoodsManager { public String LoadGoods();
}

View Code

4、创建测试服务实现类（GoodsManager）

package business; import com.alibaba.dubbo.rpc.RpcContext; import ibusiness.IGoodsManager; public class GoodsManager implements IGoodsManager { @Override public String LoadGoods() { return "可口可乐，雀巢咖啡  " + "/n response form provider: " + RpcContext.getContext().getLocalAddress();
    }
}

View Code

5、通过spring 集成并配置dubbo测试服务，同时指定registry的zookeeper服务器地址。

xml version="1.0" encoding="UTF-8"?> <beans xmlns=""  xmlns:xsi=""  xmlns:dubbo=""  xsi:schemaLocation="   /spring-beans.xsd      /dubbo.xsd   ">    <dubbo:application name="demo-dubbo-provider-app" />    <dubbo:registry address="zookeeper://10.129.221.64:2181" />    <dubbo:protocol name="dubbo" port="30001" />    <dubbo:service interface="ibusiness.IGoodsManager" ref="goodsService" />    <bean id="goodsService" class="business.GoodsManager" /> beans> 

6、编写控制台程序启动spring容器，编译并打包Provider.jar

package dubboProvider; import org.springframework.context.ApplicationContext; import org.springframework.context.support.ClassPathXmlApplicationContext; public class DubboProvider {  public static void main(String[] args) {   try {    // 初始化Spring    ApplicationContext ctx = new ClassPathXmlApplicationContext("dubboProvider\\dubboProvider.xml");    System.out.println("dubbo provider is running...");    System.in.read();   } catch (Exception ex) {    ex.printStackTrace();   }  }
} 

View Code

五、服务消费方开发（Consumer）

1、通过spring 配置指定registry的zookeeper地址，实现对dubbo远程服务的调用

xml version="1.0" encoding="UTF-8"?> <beans xmlns=""  xmlns:xsi="" xmlns:dubbo=""  xsi:schemaLocation="  /spring-beans.xsd    /dubbo.xsd">  <dubbo:application name="demo-dubbo-consumer-app" />  <dubbo:registry address="zookeeper://10.129.221.64:2181" />    <dubbo:reference id="goodsService" interface="ibusiness.IGoodsManager" /> beans> 

2、编写调用测试客户端代码，从容器中获取远程bean并调用。

package dubboConsumer; import ibusiness.IGoodsManager; import org.springframework.context.ApplicationContext; import org.springframework.context.support.ClassPathXmlApplicationContext; public class DubboConsumer { public static void main(String[] args) {  // 初始化Spring  ApplicationContext ctx = new ClassPathXmlApplicationContext("dubboConsumer\\dubboConsumer.xml");  IGoodsManager goodsManager = (IGoodsManager) ctx.getBean("goodsService"); // 获取远程服务代理  String goodsStr = goodsManager.LoadGoods() ; // 执行远程方法  System.out.println(goodsStr);
    } 
} 

View Code

3、 分别启动 provider、consumer 的测试程序发布和调用服务，即简单实现了基于 zookeeper registry 的 dubbo 远程过程调用。

六、More about dubbo

以上简单的展示dubbo的RPC服务能力，更多关于dubbo的集群管理、监控、负载均衡、高可用性、可扩展、伸缩性等相关内容请期待后续博文。