SpringBoot 整合RabbitMQ 之延迟队列实验

喜欢ヅ旅行 2024-02-19 09:44 6阅读 0赞

## 系列文章目录 ##

[第一章 Java线程池技术应用][Java]  
[第二章 CountDownLatch和Semaphone的应用][CountDownLatch_Semaphone]  
[第三章 Spring Cloud 简介][Spring Cloud]  
[第四章 Spring Cloud Netflix 之 Eureka][Spring Cloud Netflix _ Eureka]  
[第五章 Spring Cloud Netflix 之 Ribbon][Spring Cloud Netflix _ Ribbon]  
[第六章 Spring Cloud 之 OpenFeign][Spring Cloud _ OpenFeign]  
[第七章 Spring Cloud 之 GateWay][Spring Cloud _ GateWay]  
[第八章 Spring Cloud Netflix 之 Hystrix][Spring Cloud Netflix _ Hystrix]  
[第九章 代码管理gitlab 使用][_gitlab]  
[第十章 SpringCloud Alibaba 之 Nacos discovery][SpringCloud Alibaba _ Nacos discovery]  
[第十一章 SpringCloud Alibaba 之 Nacos Config][SpringCloud Alibaba _ Nacos discovery]  
[第十二章 Spring Cloud Alibaba 之 Sentinel][Spring Cloud Alibaba _ Sentinel]  
[第十三章 JWT][JWT]  
[第十四章 RabbitMQ应用][RabbitMQ]  
[第十五章 RabbitMQ 延迟队列][RabbitMQ 1]

![在这里插入图片描述][8692ff151c5c4afdaf8f38ad8911ceef.png]

--------------------

### 文章目录  系列文章目录  @\[TOC\](文章目录)  前言  1、RabbitMQ延迟队列  1.1、方式1：RabbitMQ通过死信机制来实现延迟队列的功能  1.2、方式二：安装延迟队列插件  1.2.1、安装延迟队列插件：  2、消息确认机制  2.1、生产确认  2.2、消费确认 ###

## 前言 ##

实际业务中，例如秒杀系统，秒杀商品成功会有截止时间，这时需要用到RabbitMQ延迟服务。

## 1、RabbitMQ延迟队列 ##

### 1.1、方式1：RabbitMQ通过死信机制来实现延迟队列的功能 ###

*  TTL ，即 Time-To-Live，存活时间，消息和队列都可以设置存活时间
 *  Dead Letter，即死信，若给消息设置了存活时间，当超过存活时间后消息还没有被消费，则该消息变成了死信
 *  Dead Letter Exchanges（DLX），即死信交换机
 *  Dead Letter Routing Key （DLK），死信路由键

/***********************延迟队列*************************/
    //创建立即消费队列
    @Bean
    public Queue immediateQueue(){
        
        return new Queue("immediateQueue");
    }
    //创建立即消费交换机
    @Bean
    public DirectExchange immediateExchange(){
        
        return new DirectExchange("immediateExchange");
    }
    @Bean
    public Binding bindingImmediate(@Qualifier("immediateQueue") Queue queue,@Qualifier("immediateExchange") DirectExchange directExchange){
        
        return BindingBuilder.bind(queue).to(directExchange).with("immediateRoutingKey");
    }
    
    //创建延迟队列
    @Bean
    public Queue delayQueue(){
        
        Map<String,Object> params = new HashMap<>();
        //死信队列转发的死信转发到立即处理信息的交换机
        params.put("x-dead-letter-exchange","immediateExchange");
        //死信转化携带的routing-key
        params.put("x-dead-letter-routing-key","immediateRoutingKey");
        //设置消息过期时间，单位：毫秒
        params.put("x-message-ttl",60 * 1000);
        return new Queue("delayQueue",true,false,false,params);
    }
    
    @Bean
    public DirectExchange delayExchange(){
        
        return new DirectExchange("delayExchange");
    }
    
    @Bean
    public Binding bindingDelay(@Qualifier("delayQueue") Queue queue,@Qualifier("delayExchange") DirectExchange directExchange){
        
        return BindingBuilder.bind(queue).to(directExchange).with("delayRoutingKey");
    }

@Test
    public void sendDelay(){
        
        this.rabbitTemplate.convertAndSend("delayExchange","delayRoutingKey","Hello world topic");
    }

### 1.2、方式二：安装延迟队列插件 ###

#### 1.2.1、安装延迟队列插件： ####

https://github.com/rabbitmq/rabbitmq-delayed-message-exchange/releases/download/3.11.1/rabbitmq\_delayed\_message\_exchange-3.11.1.ez  
下载解压，到plugins目录，执行以下的命令：

> rabbitmq-plugins enable rabbitmq\_delayed\_message\_exchange

/**************延迟队列一个单一queue*******************/
    @Bean
    public Queue delayNewQueue(){
        
        return new Queue("delayNewQueue");
    }
    @Bean
    public CustomExchange delayNewExchange(){
        
        Map<String, Object> args = new HashMap<>();
        // 设置类型，可以为fanout、direct、topic
        args.put("x-delayed-type", "direct");
        return new CustomExchange("delayNewExchange","x-delayed-message", true,false,args);
    }
    @Bean
    public Binding bindingNewDelay(@Qualifier("delayNewQueue") Queue queue,@Qualifier("delayNewExchange") CustomExchange customExchange){
        
        return BindingBuilder.bind(queue).to(customExchange).with("delayNewRoutingKey").noargs();
    }

@Test
    public void sendDelay() {
        
        //生产端写完了
        UserInfo userInfo = new UserInfo();
        userInfo.setPassword("13432432");
        userInfo.setUserAccount("kelvin");
        this.rabbitTemplate.convertAndSend("delayNewExchange", "delayNewRoutingKey", userInfo
                , a -> {
        
                    //单位毫秒
                    a.getMessageProperties().setDelay(30000);
                    return a;
                });
    }

## 2、消息确认机制 ##

消息确认分为两部分： 生产确认 和 消费确认。

生产确认： 生产者生产消息后，将消息发送到交换机，触发确认回调；交换机将消息转发到绑定队列，若失败则触发返回回调。  
消费确认： 默认情况下消息被消费者从队列中获取后即发送确认，不管消费者处理消息时是否失败，不需要额外代码，但是不能保证消息被正确消费。我们增加手动确认，则需要代码中明确进行消息确认。

### 2.1、生产确认 ###

@Bean
    public RabbitTemplate getTemplate(ConnectionFactory connectionFactory){
        
        RabbitTemplate template = new RabbitTemplate(connectionFactory);
        //消息发送到交换器Exchange后触发回调
        template.setConfirmCallback(new RabbitTemplate.ConfirmCallback() {
        
            @Override
            public void confirm(CorrelationData correlationData, boolean ack, String cause) {
        
                //  可以进行消息入库操作
                log.info("消息唯一标识 correlationData = {}", correlationData);
                log.info("确认结果 ack = {}", ack);
                log.info("失败原因 cause = {}", cause);
            }
        });
    
        // 配置这个，下面的ReturnCallback 才会起作用
        template.setMandatory(true);
        // 如果消息从交换器发送到对应队列失败时触发（比如 根据发送消息时指定的routingKey找不到队列时会触发）
        template.setReturnsCallback(new RabbitTemplate.ReturnsCallback() {
        
            @Override
            public void returnedMessage(ReturnedMessage returnedMessage) {
        
                //  可以进行消息入库操作
                log.info("消息主体 message = {}", returnedMessage.getMessage());
                log.info("回复码 replyCode = {}", returnedMessage.getReplyCode());
                log.info("回复描述 replyText = {}", returnedMessage.getReplyText());
                log.info("交换机名字 exchange = {}", returnedMessage.getExchange());
                log.info("路由键 routingKey = {}", returnedMessage.getRoutingKey());
    
            }
        });
        return template;
    }

spring:
      cloud:
        nacos:
          discovery:
            server-addr: localhost:8848
      application:
        name: drp-user-service  #微服务名称
      datasource:
        username: root
        password: root
        url: jdbc:mysql://127.0.0.1:3306/drp
        driver-class-name: com.mysql.cj.jdbc.Driver
      rabbitmq:
        host: 127.0.0.1
        port: 5672
        username: root
        password: root
        virtual-host: root_vh
        # 确认消息已发送到交换机（Exchange）
        publisher-confirm-type: correlated
        # 确认消息已发送到队列
        publisher-returns: true
        listener:
          simple:
            acknowledge-mode: manual # 开启消息消费手动确认
            retry:
              enabled: true

### 2.2、消费确认 ###

@RabbitHandler
    public void process(UserInfo data, Message message, Channel channel){
        
        log.info("收到directQueue队列信息：" + data);
        long deliveryTag = message.getMessageProperties().getDeliveryTag();
        try {
        
            //成功消费确认
            channel.basicAck(deliveryTag,true);
            log.info("消费成功确认完毕。。。。。");
        } catch (IOException e) {
        
            log.error("确认消息时抛出异常 ", e);        
            // 重新确认，成功确认消息
            try {
        
                Thread.sleep(50);
                channel.basicAck(deliveryTag, true);
            } catch (IOException | InterruptedException e1) {
        
                log.error("确认消息时抛出异常 ", e);
                // 可以考虑入库
            }
        }catch (Exception e){
        
            log.error("业务处理失败", e);
            try {
        
                // 失败确认
                channel.basicNack(deliveryTag, false, false);
            } catch (IOException e1) {
        
                log.error("消息失败确认失败", e1);
            }
        }
    }

[Java]: https://blog.csdn.net/s445320/article/details/132870714
[CountDownLatch_Semaphone]: https://blog.csdn.net/s445320/article/details/132912126
[Spring Cloud]: https://blog.csdn.net/s445320/article/details/132939152
[Spring Cloud Netflix _ Eureka]: https://blog.csdn.net/s445320/article/details/133024912
[Spring Cloud Netflix _ Ribbon]: https://blog.csdn.net/s445320/article/details/133156575
[Spring Cloud _ OpenFeign]: https://blog.csdn.net/s445320/article/details/130624259
[Spring Cloud _ GateWay]: https://blog.csdn.net/s445320/article/details/133612365
[Spring Cloud Netflix _ Hystrix]: https://blog.csdn.net/s445320/article/details/133759729
[_gitlab]: https://blog.csdn.net/s445320/article/details/133842594
[SpringCloud Alibaba _ Nacos discovery]: https://blog.csdn.net/s445320/article/details/133977189
[Spring Cloud Alibaba _ Sentinel]: https://blog.csdn.net/s445320/article/details/134149339
[JWT]: https://blog.csdn.net/s445320/article/details/134197122
[RabbitMQ]: https://blog.csdn.net/s445320/article/details/134228181
[RabbitMQ 1]: https://blog.csdn.net/s445320/article/details/130823776
[8692ff151c5c4afdaf8f38ad8911ceef.png]: https://image.dandelioncloud.cn/pgy_files/images/2024/02/19/297bcd0c28d1480e80ea6432e223c6b2.png