当前位置：首页 > news >正文

【RabbitMQ】死信队列

news 来源：原创 2025/9/18 14:59:44

1.概述

死信，顾名思义就是无法被消费的消息，也就是没有被传到消费者的消息，或者即使传到了也没有被消费。当然有死信就有死信队列。死信队列就是用来存储死信的。

它的应用场景就是保证订单业务的消息数据不丢失，当消息消费发生异常时，将消息投入死信队列中。还有比如说：用户在商城下单成功并点击去支付后在指定时间未支付时自动失效。

死信的来源有3种：

1.消息TTL过期（TTL就是过期时间）

2.队列达到最大长度(也就是队列装满了消息)

3.消息被拒绝（(basic.reject 或 basic.nack)）并且requeue=false

解释一下这里的requeue=false

在调用basic.reject或basic.nack 方法时，都有一个参数用于决定是否将被拒绝的消息重新放回原队列。当requeue设置为false 时，被拒绝的消息不会再回到原队列。这种情况下，若队列配置了死信交换机（通过x-dead-letter-exchange参数设置）等相关死信处理机制，消息就会成为死信消息，被转发到死信交换机，再由死信交换机根据路由键转发到对应的死信队列。如果没有配置死信相关机制，消息就会被直接丢弃

2.代码实现

在编写代码之前，先看看整个流程图

生产者生产消息发送到普通交换机中，交换机根据routing key将消息转发给相应的普通队列。当普通队列中的消息由于某些原因变成了死信消息，会把死信消息转发到相应的死信交换机中，死信交换机同样会根据routing key转发给相应的死信队列，然后可以安排专门的消费者去消费死信队列中的死信消息

下面分3种情况来讲，就是按照上述3种变成死信消息的情况

2.1消费超时

生产者

public class DeadProducer {public static void main(String[] args) throws Exception {Connection connection = RabbitMQUtils.getConnection();Channel channel = connection.createChannel();channel.exchangeDeclare("exchange-normal", "direct");//模拟消息超时，超过10秒钟，消息进入死信AMQP.BasicProperties properties = new AMQP.BasicProperties().builder().expiration("10000").build();for (int i = 1; i <= 10; i++) {channel.basicPublish("exchange-normal", "zhangsan", properties, ("message" + i).getBytes());}channel.close();connection.close();}
}

消费者(处理正常消息)

import com.rabbitmq.client.*;import java.io.IOException;
import java.util.HashMap;
import java.util.Map;public class DeadConsumer1 {public static void main(String[] args) throws Exception {Connection connection = RabbitMQUtils.getConnection();//声明channelChannel channel = connection.createChannel();//声明死信和普通交换机 类型都为directchannel.exchangeDeclare("exchange-normal", "direct");channel.exchangeDeclare("exchange-dead", "direct");//声明普通队列Map<String, Object> params = new HashMap<>();//正常队列设置死信交换机 参数 key 是固定值params.put("x-dead-letter-exchange", "exchange-dead");//正常队列设置死信 routing-key 参数 key 是固定值params.put("x-dead-letter-routing-key", "lisi");String normalQueue = "normal-queue";channel.queueDeclare(normalQueue, false, false, false, params);//生成死信队列String deadQueue = "dead-queue";channel.queueDeclare(deadQueue, false, false, false, null);//绑定死信channel.queueBind(deadQueue, "exchange-dead", "lisi");//绑定普通队列和交换机channel.queueBind(normalQueue, "exchange-normal", "zhangsan");//消费消息channel.basicConsume(normalQueue, true, new DefaultConsumer(channel) {@Overridepublic void handleDelivery(String consumerTag, Envelope envelope, AMQP.BasicProperties properties, byte[] body) throws IOException {System.out.println("消费的消息是:" + new String(body));}});}
}

消费者(处理死信消息)

public class DeadConsumer2 {public static void main(String[] args) throws Exception {Connection connection = RabbitMQUtils.getConnection();//声明channelChannel channel = connection.createChannel();channel.exchangeDeclare("exchange-dead", "direct");String deadQueue = "dead-queue";channel.queueDeclare(deadQueue, false, false, false, null);//绑定死信channel.queueBind(deadQueue, "exchange-dead", "lisi");//消费消息channel.basicConsume(deadQueue, true, new DefaultConsumer(channel) {@Overridepublic void handleDelivery(String consumerTag, Envelope envelope, AMQP.BasicProperties properties, byte[] body) throws IOException {System.out.println("消费的消息是:" + new String(body));}});}
}

2.2达到队列最大长度

生产者

public class DeadProducer {public static void main(String[] args) throws Exception {Connection connection = RabbitMQUtils.getConnection();Channel channel = connection.createChannel();channel.exchangeDeclare("exchange-normal", "direct");//模拟消息超时，超过10秒钟，消息进入死信//AMQP.BasicProperties properties = new AMQP.BasicProperties().builder().expiration("10000").build();for (int i = 1; i <= 10; i++) {channel.basicPublish("exchange-normal", "zhangsan", null, ("message" + i).getBytes());}channel.close();connection.close();}
}

消费者(消费正常消息)

import com.rabbitmq.client.*;
import java.io.IOException;
import java.util.HashMap;
import java.util.Map;public class DeadConsumer1 {public static void main(String[] args) throws Exception {Connection connection = RabbitMQUtils.getConnection();//声明channelChannel channel = connection.createChannel();//声明死信和普通交换机 类型都为directchannel.exchangeDeclare("exchange-normal", "direct");channel.exchangeDeclare("exchange-dead", "direct");//声明普通队列Map<String, Object> params = new HashMap<>();//设置正常队列长度限制 key是固定值params.put("x-max-length", 6);params.put("x-dead-letter-exchange", "exchange-dead");//正常队列设置死信 routing-key 参数 key 是固定值params.put("x-dead-letter-routing-key", "lisi");String normalQueue = "normal-queue";channel.queueDeclare(normalQueue, false, false, false, params);//生成死信队列String deadQueue = "dead-queue";channel.queueDeclare(deadQueue, false, false, false, null);//绑定死信channel.queueBind(deadQueue, "exchange-dead", "lisi");//绑定普通队列和交换机channel.queueBind(normalQueue, "exchange-normal", "zhangsan");//消费消息channel.basicConsume(normalQueue, true, new DefaultConsumer(channel) {@Overridepublic void handleDelivery(String consumerTag, Envelope envelope, AMQP.BasicProperties properties, byte[] body) throws IOException {System.out.println("消费的消息是:" + new String(body));}});}
}

消费死信消息的就不重复写了，跟之前的一样。

2.3消息消费被拒绝

生产者和消费死信消息的消费者都是一样的，只需要改一下消费正常消息的消费者的代码

import com.rabbitmq.client.*;
import java.io.IOException;
import java.util.HashMap;
import java.util.Map;public class DeadConsumer1 {public static void main(String[] args) throws Exception {Connection connection = RabbitMQUtils.getConnection();//声明channelfinal Channel channel = connection.createChannel();//声明死信和普通交换机 类型都为directchannel.exchangeDeclare("exchange-normal", "direct");channel.exchangeDeclare("exchange-dead", "direct");//声明普通队列Map<String, Object> params = new HashMap<>();//设置正常队列长度限制 key是固定值//params.put("x-max-length", 6);params.put("x-dead-letter-exchange", "exchange-dead");//正常队列设置死信 routing-key 参数 key 是固定值params.put("x-dead-letter-routing-key", "lisi");String normalQueue = "normal-queue";channel.queueDeclare(normalQueue, false, false, false, params);//生成死信队列String deadQueue = "dead-queue";channel.queueDeclare(deadQueue, false, false, false, null);//绑定死信channel.queueBind(deadQueue, "exchange-dead", "lisi");//绑定普通队列和交换机channel.queueBind(normalQueue, "exchange-normal", "zhangsan");//消费消息channel.basicConsume(normalQueue, false, new DefaultConsumer(channel) {@Overridepublic void handleDelivery(String consumerTag, Envelope envelope, AMQP.BasicProperties properties, byte[] body) throws IOException {String message = new String(body);if (message.equals("message5")) {System.out.println("Consumer01接收到消息" + message + "并拒绝签收该消息");//拒绝消费该消息channel.basicReject(envelope.getDeliveryTag(), false);} else {System.out.println("消费的消息是:" + message);channel.basicAck(envelope.getDeliveryTag(), false);}}});}
}

用equals进行匹配，如果是指定的消息，就拒绝消费。

1.概述

2.代码实现

2.1消费超时

2.2达到队列最大长度

2.3消息消费被拒绝

相关文章：