【4】 -Routing

最后更新于:2022-04-01 15:58:44

#  RabbitMQ入门教程 For Java【4】 - [Routing](http://www.rabbitmq.com/tutorials/tutorial-four-python.html) ### **我的开发环境:** 操作系统:**Windows7 64bit** 开发环境:**JDK 1.7 - 1.7.0_55** 开发工具:**Eclipse Kepler SR2** RabbitMQ版本: **3.6.0** Elang版本:**erl7.2.1** 关于Windows7下安装RabbitMQ的教程请先在网上找一下,有空我再补安装教程。 ### 源码地址 https://github.com/chwshuang/rabbitmq.git ### 消息路由         上一章教程中我们建立了一个简单的日志记录系统,能够将消息广播到多个消费者。本章,我们将添加一个新功能,类似订阅消息的子集。例如:我们只接收日志文件中ERROR类型的日志。 ### 绑定关系         在之前的例子中也使用了类似的方式: ~~~ channel.queueBind(queueName, EXCHANGE_NAME, ""); ~~~         绑定是交换器和队列之间的一种关系,用户微博,微信的例子可以简单的理解为关注,就是队列(某屌丝)对交换器(女神)非常感兴趣,关注了她,以后女神发的每条微博,屌丝都能看到。         绑定可以使用routingkey这个参数,是为了避免所有的消息都使用同一个路由线索带来的麻烦。为了区分路由规则,我们创建创建一个唯一的路由线索。 ~~~ channel.queueBind(queueName, EXCHANGE_NAME, "black"); ~~~         绑定关系中使用的路由关键字【routingkey】是否有效取决于交换器的类型。如果交换器是分发【fanout】类型,就会忽略路由关键字【routingkey】的作用。 ### 直连类型交换器         上一章的例子是通过分发【fanout】类型的交换器【logs】广播日志信息,现在我们将日志分debug、info、warn、error这几种基本的级别,实际在生产环境中,避免磁盘空间浪费,应用只会将error级别的日志打印出来。而分发【fanout】类型的交换器会将所有基本的日志都发送出来,如果我们想只接收某一级别的日志信息,就需要使用直连【direct】类型的交换器了, 下面的图中,队列1通过ERROR这个routingkey绑定到E交换器,队列2通过WARN和INFO绑定到E交换器,E交换器的类型是直连【direct】的,如果生产者【P】发出ERROR的日志,只会有队列1会收到,如果生产者【P】发出INFO和WARN的日志,只有队列2会收到,如果生产者【P】发出DEBUG级别的日志,队列1和队列2都会忽略它。 ![](https://docs.gechiui.com/gc-content/uploads/sites/kancloud/2016-03-01_56d507d75d06e.jpg) ### 多重绑定         我们允许多个队列以相同的路由关键字绑定到同一个交换器中,可以看到,交换器虽然是直连类型,但是绑定后的效果却跟分发类型的交换器类似,相同的是队列1和队列2都会收到同一条来自交换器的消息。         他们的区别:分发模式下,队列1、队列2会收到所有级别(除ERROR级别以外)的消息,而直连模式下,他们仅仅只会收到ERROR关键字类型的消息。 ![](https://docs.gechiui.com/gc-content/uploads/sites/kancloud/2016-03-01_56d507d770c51.jpg) ### 发送日志消息         我们还是用日志系统进行讲解,现在我们用日志的级别来作为路由关键字【routingkey】,这样,消费者端就可以按照他关心的日志级别进行接收,我们先看看如何发送日志:         先声明交换器 ~~~ <span style="font-size: 18px;">        </span>channel.exchangeDeclare(EXCHANGE_NAME, "direct"); ~~~         然后发送消息到交换器 ~~~ for (String severity : routingKeys) { channel.queueBind(queueName, EXCHANGE_NAME, severity); System.out.println("ReceiveLogsDirect1 exchange:"+EXCHANGE_NAME+", queue:"+queueName+", BindRoutingKey:" + severity); } ~~~ ### 订阅消息         我们先获取一个随机的队列名称,然后根据多个路由关键字【routingkey】将队列和交换器绑定起来: ~~~ String queueName = channel.queueDeclare().getQueue(); for(String severity : argv){ channel.queueBind(queueName, EXCHANGE_NAME, severity); } ~~~ ### 项目说明 ### 流程图 ![](https://docs.gechiui.com/gc-content/uploads/sites/kancloud/2016-03-01_56d507d77fafc.jpg) ### 包图 ![](https://docs.gechiui.com/gc-content/uploads/sites/kancloud/2016-03-01_56d507d793b2e.jpg) ### 代码 RoutingSendDirect.java ~~~ import com.rabbitmq.client.ConnectionFactory; import com.rabbitmq.client.Connection; import com.rabbitmq.client.Channel; /** * @author hushuang * */ public class RoutingSendDirect { private static final String EXCHANGE_NAME = "direct_logs"; // 路由关键字 private static final String[] routingKeys = new String[]{"info" ,"warning", "error"}; public static void main(String[] argv) throws Exception { ConnectionFactory factory = new ConnectionFactory(); factory.setHost("localhost"); Connection connection = factory.newConnection(); Channel channel = connection.createChannel(); // 声明交换器 channel.exchangeDeclare(EXCHANGE_NAME, "direct"); // 发送消息 for(String severity :routingKeys){ String message = "Send the message level:" + severity; channel.basicPublish(EXCHANGE_NAME, severity, null, message.getBytes()); System.out.println(" [x] Sent '" + severity + "':'" + message + "'"); } channel.close(); connection.close(); } } ~~~ ReceiveLogsDirect1.java和ReceiveLogsDirect2.java ~~~ import com.rabbitmq.client.*; import java.io.IOException; public class ReceiveLogsDirect1 { // 交换器名称 private static final String EXCHANGE_NAME = "direct_logs"; // 路由关键字 private static final String[] routingKeys = new String[]{"info" ,"warning", "error"}; public static void main(String[] argv) throws Exception { ConnectionFactory factory = new ConnectionFactory(); factory.setHost("localhost"); Connection connection = factory.newConnection(); Channel channel = connection.createChannel(); // 声明交换器 channel.exchangeDeclare(EXCHANGE_NAME, "direct"); // 获取匿名队列名称 String queueName = channel.queueDeclare().getQueue(); // 根据路由关键字进行多重绑定 for (String severity : routingKeys) { channel.queueBind(queueName, EXCHANGE_NAME, severity); System.out.println("ReceiveLogsDirect1 exchange:"+EXCHANGE_NAME+", queue:"+queueName+", BindRoutingKey:" + severity); } System.out.println("ReceiveLogsDirect1 [*] Waiting for messages. To exit press CTRL+C"); Consumer consumer = new DefaultConsumer(channel) { @Override public void handleDelivery(String consumerTag, Envelope envelope, AMQP.BasicProperties properties, byte[] body) throws IOException { String message = new String(body, "UTF-8"); System.out.println(" [x] Received '" + envelope.getRoutingKey() + "':'" + message + "'"); } }; channel.basicConsume(queueName, true, consumer); } } ~~~ ~~~ import com.rabbitmq.client.*; import java.io.IOException; public class ReceiveLogsDirect2 { // 交换器名称 private static final String EXCHANGE_NAME = "direct_logs"; // 路由关键字 private static final String[] routingKeys = new String[]{"error"}; public static void main(String[] argv) throws Exception { ConnectionFactory factory = new ConnectionFactory(); factory.setHost("localhost"); Connection connection = factory.newConnection(); Channel channel = connection.createChannel(); // 声明交换器 channel.exchangeDeclare(EXCHANGE_NAME, "direct"); // 获取匿名队列名称 String queueName = channel.queueDeclare().getQueue(); // 根据路由关键字进行多重绑定 for (String severity : routingKeys) { channel.queueBind(queueName, EXCHANGE_NAME, severity); System.out.println("ReceiveLogsDirect1 exchange:"+EXCHANGE_NAME+", queue:"+queueName+", BindRoutingKey:" + severity); } System.out.println("ReceiveLogsDirect1 [*] Waiting for messages. To exit press CTRL+C"); Consumer consumer = new DefaultConsumer(channel) { @Override public void handleDelivery(String consumerTag, Envelope envelope, AMQP.BasicProperties properties, byte[] body) throws IOException { String message = new String(body, "UTF-8"); System.out.println(" [x] Received '" + envelope.getRoutingKey() + "':'" + message + "'"); } }; channel.basicConsume(queueName, true, consumer); } } ~~~ 先运行ReceiveLogsDirect1.java和ReceiveLogsDirect2.java 查看日志,我们发现RabbitMQ中已经创建了direct_logs的交换器,以及amq.gen-dVUpkqxmladY3Jg1upDsDQ 和amq.gen-skrmBAlYKSDzELKtVg_zFw这两个临时队列, ~~~ ReceiveLogsDirect1 exchange:direct_logs, queue:amq.gen-skrmBAlYKSDzELKtVg_zFw, BindRoutingKey:info ReceiveLogsDirect1 exchange:direct_logs, queue:amq.gen-skrmBAlYKSDzELKtVg_zFw, BindRoutingKey:warning ReceiveLogsDirect1 exchange:direct_logs, queue:amq.gen-skrmBAlYKSDzELKtVg_zFw, BindRoutingKey:error ReceiveLogsDirect1 [*] Waiting for messages. To exit press CTRL+C ~~~ ~~~ ReceiveLogsDirect2 exchange:direct_logs, queue:amq.gen-dVUpkqxmladY3Jg1upDsDQ, BindRoutingKey:error ReceiveLogsDirect2 [*] Waiting for messages. To exit press CTRL+C ~~~ 运行RoutingSendDirect.java发送消息: ### 运行结果 查看日志: RoutingSendDirect.java ~~~ [x] Sent 'info':'Send the message level:info' [x] Sent 'warning':'Send the message level:warning' [x] Sent 'error':'Send the message level:error' ~~~ ReceiveLogsDirect1.java ~~~ [x] Received 'info':'Send the message level:info' [x] Received 'warning':'Send the message level:warning' [x] Received 'error':'Send the message level:error' ~~~ ReceiveLogsDirect2.java ~~~ [x] Received 'error':'Send the message level:error' ~~~ 我们看到,队列1收到了所有的消息,队列2只收到了error级别的消息。这与我们的预期一样。 下一阶段我们可以进入第五章-主题的学习了。 本教程所有文章: [RabbitMQ入门教程 For Java【1】 - Hello World](http://blog.csdn.net/chwshuang/article/details/50521708)  - 你好世界!  [RabbitMQ入门教程 For Java【2】 - Work Queues](http://blog.csdn.net/chwshuang/article/details/50506284)  - 工作队列 [RabbitMQ入门教程 For Java【3】 - Publish/Subscribe](http://blog.csdn.net/chwshuang/article/details/50512057) - 发布/订阅 [RabbitMQ入门教程 For Java【4】 - Routing](http://blog.csdn.net/chwshuang/article/details/50505060) -  消息路由 [RabbitMQ入门教程 For Java【5】 - Topic](http://blog.csdn.net/chwshuang/article/details/50516904)  -  模糊匹配 [RabbitMQ入门教程 For Java【6】 - Remote procedure call (RPC)](http://blog.csdn.net/chwshuang/article/details/50518570) - 远程调用
';