代码仓库:github:https://github.com/stopping5/RabbitMq-Operation-Record.git
本代码示例需要引入rabbitmq依赖
<!-- rabbitmq依赖客户端-->
<dependency>
<groupId>com.rabbitmq</groupId>
<artifactId>amqp-client</artifactId>
<version>5.8.0</version>
</dependency>
一、什么是发布确认
Publisher confirms are a RabbitMQ extension to implement reliable publishing. When publisher confirms are enabled on a channel, messages the client publishes are confirmed asynchronously by the broker, meaning they have been taken care of on the server side.
发布订阅即使生产者发送消息到broker时,broker可以通过同步或者异步的方式通知消息生产者是否成功。
二、Channel启动消息确认
Channel channel = connection.createChannel();
channel.confirmSelect();
三、消息发布确认
1. 单条同步确认[耗性能不推荐]
/**
* 单独确认
* 消息发送耗时:14492ms
* */
public static void simpleConfirmMessage() throws IOException, InterruptedException {
Connection connection = RabbitMqUtil.getConnection();
Channel channel = connection.createChannel();
channel.exchangeDeclare(RabbitMQConfig.ACK_EXCHANGE, BuiltinExchangeType.DIRECT);
//声明队列关闭了自动确认
channel.queueDeclare(RabbitMQConfig.ACK_QUEUE,false,false,false,null);
channel.queueBind(RabbitMQConfig.ACK_QUEUE,RabbitMQConfig.ACK_EXCHANGE,"ACK");
//开启消息确认
channel.confirmSelect();
long begin = System.currentTimeMillis();
for (int i = 0; i < PUBLISH_COUNT; i++) {
String message = i + "";
channel.basicPublish(RabbitMQConfig.ACK_EXCHANGE,"ACK",null,message.getBytes(StandardCharsets.UTF_8));
boolean flag = channel.waitForConfirms();
if (!flag){
System.out.println("消息发送失败");
}
}
long end = System.currentTimeMillis();
System.out.println("消息发送耗时:"+(end-begin)+"ms");
}
或者可以使用Channel#waitForConfirmsOrDie(long)消息如果在限定时间内没有答复则抛出异常
Channel#waitForConfirmsOrDie(long) method. The method returns as soon as the message has been confirmed. If the message is not confirmed within the timeout or if it is nack-ed (meaning the broker could not take care of it for some reason), the method will throw an exception.
2. 批量确认同步确认
使用方式和单条发布确认一样,只不是对消息发布确认频率修改成批量确认模式,若确认失败只能知道一批消息有消息发送失败,无法精确到某一条。
/**
* 批量发布确认
* 批量确认消息发送耗时:268ms
* */
public static void batchConfirmMessage() throws IOException, InterruptedException {
Connection connection = RabbitMqUtil.getConnection();
Channel channel = connection.createChannel();
channel.exchangeDeclare(RabbitMQConfig.ACK_EXCHANGE, BuiltinExchangeType.DIRECT);
//声明队列关闭了自动确认
channel.queueDeclare(RabbitMQConfig.ACK_QUEUE,false,false,false,null);
channel.queueBind(RabbitMQConfig.ACK_QUEUE,RabbitMQConfig.ACK_EXCHANGE,"ACK");
//开启消息确认
channel.confirmSelect();
long begin = System.currentTimeMillis();
int batchSize = 100;
for (int i = 0; i < PUBLISH_COUNT; i++) {
String message = i + "";
channel.basicPublish(RabbitMQConfig.ACK_EXCHANGE,"ACK",null,message.getBytes(StandardCharsets.UTF_8));
if (i%batchSize == 0){
boolean flag = channel.waitForConfirms();
if (flag){
System.out.println("消息发送成功");
}
}
}
long end = System.currentTimeMillis();
System.out.println("批量确认消息发送耗时:"+(end-begin)+"ms");
}
3. 异步发布确认[推荐]
相对于同步确认而言,异步确认机制新增监听器体提供确认成功和确认失败的方法,复用方法即可异步的获取到消息发送到broker的结果。
/**
* 批量发布确认
* 异步确认消息发送耗时:36ms
* */
public static void asynConfirmMessage() throws IOException, InterruptedException {
Connection connection = RabbitMqUtil.getConnection();
Channel channel = connection.createChannel();
channel.exchangeDeclare(RabbitMQConfig.ACK_EXCHANGE, BuiltinExchangeType.DIRECT);
//声明队列关闭了自动确认
channel.queueDeclare(RabbitMQConfig.ACK_QUEUE,false,false,false,null);
channel.queueBind(RabbitMQConfig.ACK_QUEUE,RabbitMQConfig.ACK_EXCHANGE,"ACK");
//开启消息确认
channel.confirmSelect();
//设置异步监听回调方法
ConfirmCallback ackCallback = (deliveryTag,multipl)->{
System.out.println("消息发布确认成功:"+deliveryTag);
};
ConfirmCallback unAckCallback = (deliveryTag,multipl)->{
System.out.println("消息发布失败:"+deliveryTag);
};
channel.addConfirmListener(ackCallback,unAckCallback);
long begin = System.currentTimeMillis();
for (int i = 0; i < PUBLISH_COUNT; i++) {
String message = i + "";
channel.basicPublish(RabbitMQConfig.ACK_EXCHANGE,"ACK",null,message.getBytes(StandardCharsets.UTF_8));
}
long end = System.currentTimeMillis();
System.out.println("异步确认消息发送耗时:"+(end-begin)+"ms");
}
四、总结
1. 优缺点
- 单条同步发布确认:publishing messages individually, waiting for the confirmation synchronously: simple, but very limited throughput.
- 批量同步发布确认:publishing messages in batch, waiting for the confirmation synchronously for a batch: simple, reasonable throughput, but hard to reason about when something goes wrong.
- 异步发布确认:asynchronous handling: best performance and use of resources, good control in case of error, but can be involved to implement correctly.
2. 效率
测试结果可以看出异步确认效率是最高的,其次是批量确认,最后当然是一条一条的单条确认。
public static void main(String[] args) throws IOException, InterruptedException {
//simpleConfirmMessage();//消息发送耗时:14492ms
//batchConfirmMessage();//批量确认消息发送耗时:268ms
//asynConfirmMessage();//异步确认消息发送耗时:36ms
}