Dubbo的一次体验与分析
KathrinSuar
8年前
<p>Dubbo是什么</p> <p>Dubbo是一个分布式服务框架,致力于提供高性能和透明化的RPC远程服务调用方案,以及SOA服务治理方案;</p> <p>其核心部分包含:</p> <p>远程通讯:提供对多种基于长连接的NIO框架抽象封装,包括多种线程模型,序列化,以及“请求-响应”模式的信息交换方式。</p> <p>集群容错:提供基于接口方法的透明远程过程调用,包括多协议支持,以及软负载均衡,失败容错,地址路由,动态配置等集群支持。</p> <p>自动发现:基于注册中心目录服务,使服务消费方能动态的查找服务提供方,使地址透明,使服务提供方可以平滑增加或减少机器。</p> <p>架构图如下:来自官网</p> <p style="text-align:center"><img src="https://simg.open-open.com/show/f07431ff8daafff6906882197d395a2a.jpg"></p> <p>Provider: 暴露服务的服务提供方。</p> <p>Consumer: 调用远程服务的服务消费方。</p> <p>Registry: 服务注册与发现的注册中心。</p> <p>Monitor: 统计服务的调用次调和调用时间的监控中心。</p> <p>Container: 服务运行容器。</p> <p>与Spring和Zookeeper的集成测试</p> <p>从上面的架构图中了解到整个系统需要5个部分,不过此处为了方便测试,只提供了Provider,Consumer以及Registry三个部分;代码结构图如下所示:</p> <p><img src="https://simg.open-open.com/show/3b694f7a717b6f7041ed401d77cb106e.jpg"></p> <p>dubboTest是公共的父工程,dubboProvider和dubboConsumer分别是其子工程,分别对应了Provider和Consumer两个部分,至于Registry由Zookeeper来支持,下面详细介绍每一部分。</p> <p>1.公共maven依赖</p> <pre> <code class="language-java"><dependency> <groupId>org.springframework</groupId> <artifactId>spring-webmvc</artifactId> <version>4.3.3.RELEASE</version> </dependency> <dependency> <groupId>org.apache.zookeeper</groupId> <artifactId>zookeeper</artifactId> <version>3.4.6</version> </dependency> <dependency> <groupId>com.github.sgroschupf</groupId> <artifactId>zkclient</artifactId> <version>0.1</version> </dependency> <dependency> <groupId>com.alibaba</groupId> <artifactId>dubbo</artifactId> <version>2.5.3</version> <exclusions> <exclusion> <artifactId>spring</artifactId> <groupId>org.springframework</groupId> </exclusion> </exclusions> </dependency></code></pre> <p>主要是Spring,Zookeeper以及dubbo相关的包</p> <p>2.Provider相关介绍</p> <p>提供一个对外的接口类DemoService</p> <pre> <code class="language-java">public interface DemoService { String syncSayHello(String name); String asyncSayHello(String name); }</code></pre> <p>提供DemoService的实现类DemoServiceImpl</p> <pre> <code class="language-java">public class DemoServiceImpl implements DemoService { @Override public String syncSayHello(String name) { return "sync Hello " + name; } @Override public String asyncSayHello(String name) { return "async Hello " + name; } }</code></pre> <p>提供provider配置文件</p> <pre> <code class="language-java"><?xml version="1.0" encoding="UTF-8"?> <beans xmlns="http://www.springframework.org/schema/beans" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:dubbo="http://code.alibabatech.com/schema/dubbo" xsi:schemaLocation="http://www.springframework.org/schema/beans http://www.springframework.org/schema/beans/spring-beans.xsd http://code.alibabatech.com/schema/dubbo http://code.alibabatech.com/schema/dubbo/dubbo.xsd"> <!-- 提供方应用信息,用于计算依赖关系 --> <dubbo:application name="hello-world-app" /> <!-- 使用zookeeper注册中心暴露服务地址 --> <dubbo:registry address="zookeeper://127.0.0.1:2181" /> <!-- 用dubbo协议在20880端口暴露服务 --> <dubbo:protocol name="dubbo" port="20880" /> <!-- 声明需要暴露的服务接口 --> <dubbo:service interface="org.dubboProvider.DemoService" ref="demoService" /> <!-- 和本地bean一样实现服务 --> <bean id="demoService" class="org.dubboProvider.DemoServiceImpl" /> </beans></code></pre> <p>dubbo:registry:提供了注册中心,为了方便此处配置的是本地的zookeeper</p> <p>dubbo:service:提供了对外的服务接口</p> <p>提供Provider启动类Provider</p> <pre> <code class="language-java">import org.springframework.context.support.ClassPathXmlApplicationContext; public class Provider { public static void main(String[] args) throws Exception { ClassPathXmlApplicationContext context = new ClassPathXmlApplicationContext( new String[] { "dubbo-provider.xml" }); context.start(); System.in.read(); // 按任意键退出 } }</code></pre> <p>3.Consumer相关介绍</p> <p>Consumer配置文件dubbo-consumer.xml</p> <pre> <code class="language-java"><?xml version="1.0" encoding="UTF-8"?> <beans xmlns="http://www.springframework.org/schema/beans" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:dubbo="http://code.alibabatech.com/schema/dubbo" xsi:schemaLocation="http://www.springframework.org/schema/beans http://www.springframework.org/schema/beans/spring-beans.xsd http://code.alibabatech.com/schema/dubbo http://code.alibabatech.com/schema/dubbo/dubbo.xsd"> <!-- 消费方应用名,用于计算依赖关系,不是匹配条件,不要与提供方一样 --> <dubbo:application name="consumer-of-helloworld-app" /> <dubbo:registry address="zookeeper://127.0.0.1:2181" /> <!-- 生成远程服务代理,可以和本地bean一样使用demoService --> <dubbo:reference id="demoService" interface="org.dubboProvider.DemoService" > <dubbo:method name="syncSayHello" async="false" /> <dubbo:method name="asyncSayHello" async="true" /> </dubbo:reference> </beans></code></pre> <p>Consumer测试类Consumer</p> <pre> <code class="language-java">import java.util.concurrent.ExecutionException; import java.util.concurrent.Future; import org.dubboProvider.DemoService; import org.springframework.context.support.ClassPathXmlApplicationContext; import com.alibaba.dubbo.rpc.RpcContext; public class Consumer { public static void main(String[] args) throws InterruptedException, ExecutionException { ClassPathXmlApplicationContext context = new ClassPathXmlApplicationContext( "dubbo-consumer.xml"); context.start(); DemoService demoService = (DemoService) context.getBean("demoService"); // 获取远程服务代理 System.out.println(demoService.syncSayHello("world")); System.out.println(demoService.asyncSayHello("world")); Future<String> futrue = RpcContext.getContext().getFuture(); System.out.println(futrue.get()); } }</code></pre> <p>运行Consumer类,结果如下:</p> <pre> <code class="language-java">sync Hello world null async Hello world</code></pre> <p>调用流程简要分析</p> <p>1.启动Provider,读取配置文件</p> <p>Provider启动向Zookeeper中注册服务器的相关信息,主要的接口类RegistryService,四个实现类分别是:ZookeeperRegistry,RedisRegistry,DubboRegistry以及MulticastRegistry,这里使用的是ZookeeperRegistry,启动Provider的时候会调用doRegistry()方法,代码如下:</p> <pre> <code class="language-java">protected void doRegister(URL url) { try { zkClient.create(toUrlPath(url), url.getParameter(Constants.DYNAMIC_KEY, true)); } catch (Throwable e) { throw new RpcException("Failed to register " + url + " to zookeeper " + getUrl() + ", cause: " + e.getMessage(), e); } }</code></pre> <p>url详细信息:dubbo://192.168.67.13:20880/org.dubboProvider.DemoService?anyhost=true&application=hello-world-app&dubbo=2.5.3&interface=org.dubboProvider.DemoService&methods=syncSayHello,asyncSayHello&pid=4952&side=provider×tamp=1487671777581</p> <p>通过zkclient在Zookeeper上创建节点,为Consumer获取节点做准备。</p> <p>2.启动Consumer,读取配置文件</p> <p>监听Zookeeper中注册的服务器信息节点,通过节点信息建立和远程服务器的连接,所有的客户端都继承于AbstractClient,对应的实现类有NettyClient,MinaClient以及GrizzlyClient;3个都是基于java nio的底层通信框架,默认使用的是NettyClient,在AbstractClient的构造方法中就建立了和服务器的连接,部分代码如下:</p> <pre> <code class="language-java">public AbstractClient(URL url, ChannelHandler handler) throws RemotingException { //...省略... connect(); //...省略... }</code></pre> <p>对应的在NettyClient中进行了doConnect()</p> <pre> <code class="language-java">protected void doConnect() throws Throwable { long start = System.currentTimeMillis(); ChannelFuture future = bootstrap.connect(getConnectAddress()); //...省略... NettyClient.this.channel = newChannel; //...省略... }</code></pre> <p>NettyClient中保存了建立的连接。</p> <p>3.生成动态代理</p> <p>通过简单的接口调用就实现了远程方法的调用,其实就是dubbo帮助我们生成了一个动态代理类,所有的关于建立远程连接,消息封装编码,消息的发送以及消息的接收解码都在动态代理类里面帮我们处理了</p> <p>dubbo提供的代理工厂类都继承于AbstractProxyFactory类,对应的实现类有JdkProxyFactory和JavassistProxyFactory,默认情况下使用的JavassistProxyFactory,相应代码如下:</p> <pre> <code class="language-java">public <T> T getProxy(Invoker<T> invoker, Class<?>[] interfaces) { return (T) Proxy.getProxy(interfaces).newInstance(new InvokerInvocationHandler(invoker)); }</code></pre> <p>4.方法调用触发InvokerInvocationHandler调用invoke方法</p> <p>invoke方法代码如下:</p> <pre> <code class="language-java">public Object invoke(Object proxy, Method method, Object[] args) throws Throwable { String methodName = method.getName(); Class<?>[] parameterTypes = method.getParameterTypes(); if (method.getDeclaringClass() == Object.class) { return method.invoke(invoker, args); } if ("toString".equals(methodName) && parameterTypes.length == 0) { return invoker.toString(); } if ("hashCode".equals(methodName) && parameterTypes.length == 0) { return invoker.hashCode(); } if ("equals".equals(methodName) && parameterTypes.length == 1) { return invoker.equals(args[0]); } return invoker.invoke(new RpcInvocation(method, args)).recreate(); }</code></pre> <p>RpcInvocation用来封装了方法调用的相关参数比如:方法名,参数类型,参数列表等,可以查看相关代码:</p> <pre> <code class="language-java">public class RpcInvocation implements Invocation, Serializable { private static final long serialVersionUID = -4355285085441097045L; private String methodName; private Class<?>[] parameterTypes; private Object[] arguments; private Map<String, String> attachments; private transient Invoker<?> invoker; ......//以下省略</code></pre> <p>经过层层调用最后到达了DubboInvoker的doInvoke方法中,也是我们比较关心的类,代码如下:</p> <pre> <code class="language-java">protected Result doInvoke(final Invocation invocation) throws Throwable { RpcInvocation inv = (RpcInvocation) invocation; final String methodName = RpcUtils.getMethodName(invocation); inv.setAttachment(Constants.PATH_KEY, getUrl().getPath()); inv.setAttachment(Constants.VERSION_KEY, version); ExchangeClient currentClient; if (clients.length == 1) { currentClient = clients[0]; } else { currentClient = clients[index.getAndIncrement() % clients.length]; } try { boolean isAsync = RpcUtils.isAsync(getUrl(), invocation); boolean isOneway = RpcUtils.isOneway(getUrl(), invocation); int timeout = getUrl().getMethodParameter(methodName, Constants.TIMEOUT_KEY,Constants.DEFAULT_TIMEOUT); if (isOneway) { boolean isSent = getUrl().getMethodParameter(methodName, Constants.SENT_KEY, false); currentClient.send(inv, isSent); RpcContext.getContext().setFuture(null); return new RpcResult(); } else if (isAsync) { ResponseFuture future = currentClient.request(inv, timeout) ; RpcContext.getContext().setFuture(new FutureAdapter<Object>(future)); return new RpcResult(); } else { RpcContext.getContext().setFuture(null); return (Result) currentClient.request(inv, timeout).get(); } } catch (TimeoutException e) { throw new RpcException(RpcException.TIMEOUT_EXCEPTION, "Invoke remote method timeout. method: " + invocation.getMethodName() + ", provider: " + getUrl() + ", cause: " + e.getMessage(), e); } catch (RemotingException e) { throw new RpcException(RpcException.NETWORK_EXCEPTION, "Failed to invoke remote method: " + invocation.getMethodName() + ", provider: " + getUrl() + ", cause: " + e.getMessage(), e); } }</code></pre> <p>里面有我们比较关心的同步和异步调用:</p> <p>同步调用:调用ResponseFuture的get()方法进行等待服务器的返回</p> <p>异步调用:没有等待服务器的返回,直接将ResponseFuture放入了RpcContext.getContext()中,这也是我们需要在代码中使用Future futrue = RpcContext.getContext().getFuture();的原因。</p> <p>5.发送请求</p> <p>在第二步中已经建立了和远程服务器的连接,建立的数量和集群的服务器有关,所以客户端是一个数组,从数组中获取一个客户端连接</p> <pre> <code class="language-java">if (clients.length == 1) { currentClient = clients[0]; } else { currentClient = clients[index.getAndIncrement() % clients.length]; }</code></pre> <p>获取之后执行currentClient.request(inv, timeout),就是对服务器发送请求,进入到request方法中有如下代码:</p> <pre> <code class="language-java">public ResponseFuture request(Object request, int timeout) throws RemotingException { if (closed) { throw new RemotingException(this.getLocalAddress(), null, "Failed to send request " + request + ", cause: The channel " + this + " is closed!"); } // create request. Request req = new Request(); req.setVersion("2.0.0"); req.setTwoWay(true); req.setData(request); DefaultFuture future = new DefaultFuture(channel, req, timeout); try{ channel.send(req); }catch (RemotingException e) { future.cancel(); throw e; } return future; }</code></pre> <p>把要发送的消息封装成了一个Request对象,并且返回了DefaultFuture(继承于ResponseFuture实现了同步和异步的调用),可以看一下DefaultFuture的get()方法:</p> <pre> <code class="language-java">public Object get(int timeout) throws RemotingException { if (timeout <= 0) { timeout = Constants.DEFAULT_TIMEOUT; } if (! isDone()) { long start = System.currentTimeMillis(); lock.lock(); try { while (! isDone()) { done.await(timeout, TimeUnit.MILLISECONDS); if (isDone() || System.currentTimeMillis() - start > timeout) { break; } } } catch (InterruptedException e) { throw new RuntimeException(e); } finally { lock.unlock(); } if (! isDone()) { throw new TimeoutException(sent > 0, channel, getTimeoutMessage(false)); } } return returnFromResponse(); }</code></pre> <p>while一直在循环,直到isDone为true,或者超时了,isDone其实就是判断response是否为空:</p> <pre> <code class="language-java">public boolean isDone() { return response != null; }</code></pre> <p>6.接收消息</p> <p>DefaultFuture中的received()用来接收消息同时赋值给了Response response,这样isDone()方法可以为true了</p> <p>总结</p> <p>以上只是对dubbo粗浅的使用,以及简单的了解了一下调用的整个流程,没有太多更加的深入,主要还是项目没有实际用到;不过可以看到dubbo底层通信是基于netty,mina这种高性能的通信框架,而且通过长连接减少握手;二进制流压缩数据,比常规HTTP等短连接协议更快,这个上面没有提到,更多的可以查看编解码类DubboCodec;可以认为dubbo的性能还是相当强的。</p> <p> </p> <p>来自:http://codingo.xyz/index.php/2017/02/21/dubbo/</p> <p> </p>