今天学学消息队列RocketMQ:消息类型

        RocketMQ支持的消息类型有三种：普通消息、顺序消息、延时消息、事务消息。以下内容的代码部分都是基于rocketmq-spring-boot-starter做的。

普通消息

        普通消息是一种无序消息，消息分布在各个MessageQueue当中，以保证效率为第一使命。这种消息适用于对顺序没有要求的基础消费需求。这里的Topic和MessageQueue是多对多关系。

// 生产者
public static void main(String[] args) throws MQClientException {
    DefaultMQProducer producer = new DefaultMQProducer("rmq-group");
    producer.setNamesrvAddr("172.16.200.38:9876");
    producer.setInstanceName("producer");
    producer.start();
    try {
        for (int i = 0; i < 10; i++) {
            Thread.sleep(1000);
            Message msg = new Message("Topic-test", "testTag"
                    , (new Date() + " RocketMQ test msg " + i).getBytes());
            SendResult sendResult = producer.send(msg);
            System.out.println(sendResult.getMsgId());
            System.out.println(sendResult.getMessageQueue());
            System.out.println(sendResult.getSendStatus());
            System.out.println(sendResult.getOffsetMsgId());
            System.out.println(sendResult.getQueueOffset());
            System.out.println();
            System.out.println("============================");
        }
    } catch (Exception e) {
        e.printStackTrace();
    }
    producer.shutdown();
}

// 消费者
public static void main(String[] args) throws MQClientException {
    DefaultMQPushConsumer consumer = new DefaultMQPushConsumer("rmq-group");
    consumer.setNamesrvAddr("172.16.200.38:9876");
    consumer.setInstanceName("consumer");
    consumer.subscribe("Topic-test", "*");
    consumer.setConsumeFromWhere(ConsumeFromWhere.CONSUME_FROM_FIRST_OFFSET);
    consumer.registerMessageListener(new MessageListenerConcurrently() {
        @Override
        public ConsumeConcurrentlyStatus consumeMessage(List msgs, ConsumeConcurrentlyContext context) {
            for (MessageExt messageExt : msgs) {
                System.out.printf("%s Receive New Messages: %s %n", Thread.currentThread().getName(), new String(messageExt.getBody()));
            }

            return ConsumeConcurrentlyStatus.CONSUME_SUCCESS;
        }
    });
    //启动消费者
    consumer.start();
}

顺序消息

        当我开始对消息的时序性有要求的时候，普通消息就无法满足我们的需求了。当我们要求顺序消费的时候，我们的Topic就不能采用多对多的形式，存放在多个MessageQueue中，而是需要牺牲一部分性能，存放在一个MessageQueue中。

// 生产者
public static void main(String[] args) throws MQClientException {
    DefaultMQProducer producer = new DefaultMQProducer("rmq-group");
    producer.setNamesrvAddr("172.16.200.38:9876");
    producer.setInstanceName("producer");
    producer.start();
    try {
        for (int i = 0; i < 10; i++) {
            Thread.sleep(1000);
            Message msg = new Message("Topic-test", "testTag"
                    , (new Date() + " RocketMQ test msg " + i).getBytes());
            SendResult sendResult = producer.send(msg, (mqs, msg1, arg) -> mqs.get(0), null);

            System.out.println(sendResult.getMsgId());
            System.out.println(sendResult.getMessageQueue());
            System.out.println(sendResult.getSendStatus());
            System.out.println(sendResult.getOffsetMsgId());
            System.out.println(sendResult.getQueueOffset());
            System.out.println("RocketMQ test msg " + i);
            System.out.println("============================");
        }
    } catch (Exception e) {
        e.printStackTrace();
    }
    producer.shutdown();
}

// 消费者
public static void main(String[] args) throws MQClientException {
    DefaultMQPushConsumer consumer = new DefaultMQPushConsumer("rmq-group");
    consumer.setNamesrvAddr("172.16.200.38:9876");
    consumer.setInstanceName("consumer");
    consumer.subscribe("Topic-test", "*");
    consumer.setConsumeFromWhere(ConsumeFromWhere.CONSUME_FROM_FIRST_OFFSET);
    // 跟普通消费不同的地方，这里采用了顺序消费方法。
    consumer.registerMessageListener(new MessageListenerOrderly() {
        @Override
        public ConsumeOrderlyStatus consumeMessage(List msgs, ConsumeOrderlyContext context) {
            for (MessageExt messageExt : msgs) {
                System.out.printf("%s Receive New Messages: %s %n", Thread.currentThread().getName(), new String(messageExt.getBody()));
            }

            return ConsumeOrderlyStatus.SUCCESS;
        }
    });
    //启动消费者
    consumer.start();
}

 延时消息

        通过指定的通知时间间隔，让消息不会立刻被消费者收到。

// 生产者
public static void main(String[] args) throws MQClientException {
    DefaultMQProducer producer = new DefaultMQProducer("rmq-group");
    producer.setNamesrvAddr("172.16.200.38:9876");
    producer.setInstanceName("producer");
    producer.start();
    try {
        for (int i = 0; i < 10; i++) {
            Thread.sleep(1000);
            Message msg = new Message("Topic-test", "testTag"
                    , (new Date() + " RocketMQ test msg " + i).getBytes());
            // 指定延时等级
            // DelayTimeLevel对应的延时时间在服务端定义:rocketmq/store/src/main/java/org/apache/rocketmq/store/config/MessageStoreConfig.java
            // private String messageDelayLevel = "1s 5s 10s 30s 1m 2m 3m 4m 5m 6m 7m 8m 9m 10m 20m 30m 1h 2h";
            msg.setDelayTimeLevel(4);
            // 按毫秒延时
            msg.setDelayTimeMs(1L);
            // 按秒延时
            msg.setDelayTimeSec(1);
            SendResult sendResult = producer.send(msg);
            System.out.println(sendResult.getMsgId());
            System.out.println(sendResult.getMessageQueue());
            System.out.println(sendResult.getSendStatus());
            System.out.println(sendResult.getOffsetMsgId());
            System.out.println(sendResult.getQueueOffset());
            System.out.println();
            System.out.println("============================");
        }
    } catch (Exception e) {
        e.printStackTrace();
    }
    producer.shutdown();
}

// 消费者
public static void main(String[] args) throws MQClientException {
    DefaultMQPushConsumer consumer = new DefaultMQPushConsumer("rmq-group");
    consumer.setNamesrvAddr("172.16.200.38:9876");
    consumer.setInstanceName("consumer");
    consumer.subscribe("Topic-test", "*");
    consumer.setConsumeFromWhere(ConsumeFromWhere.CONSUME_FROM_FIRST_OFFSET);
    consumer.registerMessageListener(new MessageListenerConcurrently() {
        @Override
        public ConsumeConcurrentlyStatus consumeMessage(List msgs, ConsumeConcurrentlyContext context) {
            for (MessageExt messageExt : msgs) {
                System.out.printf("Receive New Messages: %s At %s %n", new String(messageExt.getBody()), new Date());
            }

            return ConsumeConcurrentlyStatus.CONSUME_SUCCESS;
        }
    });
    //启动消费者
    consumer.start();
}

30秒的延迟

事务消息

        事务消息是RocketMQ提供的一种高级的消息类型，支持在分布式场景下保障消息生产和本地事务的最终一致性。

事务消息的生命周期

        (1)初始化：半事务消息被生产者构建并完成初始化，待发送到RocketMQ服务端的状态。

        (2)事务待提交：半事务消息被发送到服务端，和普通消息不同，半事务消息并不会直接被服务端持久化，而是会被单独存储到事务存储系统中，等待第二阶段本地事务返回执行结果后再提交。此时消息对下游消费者不可见。

        (3)消息回滚：第二阶段如果事务执行结果明确为回滚，服务端会将半事务消息回滚，该事务消息流程终止。

        (4)提交事务待消费：第二阶段如果事务执行结果明确为提交，服务端会将半事务消息重新存储到普通存储系统中，此时消息对下游消费者可见。

        (5)消费中：消息被消费者获取，并按照消费者本地的业务逻辑进行处理的过程。此时服务端会等待消费者完成消费并提交消费结果，如果一段时间后没有收到消费者的响应，服务端会对消息进行重试处理。

        (6)消费完成：消费者完成了消费动作，并向服务端提交了消费结果，服务端标记当前消息已经被处理。

        (7)消息删除：服务端按照消息保存机制滚动清理最早的消息数据，将消息从物理文件中删除。

// 生产者
public static void main(String[] args) throws MQClientException {
    TransactionMQProducer producer = new TransactionMQProducer("rmq-group");
    TransactionListener listener = new TransactionListenerImpl();
    producer.setNamesrvAddr("172.16.200.38:9876");
    producer.setInstanceName("producer");
    producer.setTransactionListener(listener);
    producer.start();
    try {
        for (int i = 0; i < 10; i++) {
            Thread.sleep(1000);
            Message msg = new Message("Topic-test", "testTag"
                    , (new Date() + " RocketMQ test msg " + i).getBytes());
            SendResult sendResult = producer.sendMessageInTransaction(msg, null);
            System.out.println(new Date());
            System.out.println(sendResult.getMsgId());
            System.out.println(sendResult.getMessageQueue());
            System.out.println(sendResult.getSendStatus());
            System.out.println(sendResult.getOffsetMsgId());
            System.out.println(sendResult.getQueueOffset());
            System.out.println();
            System.out.println("============================");
        }
    } catch (Exception e) {
        e.printStackTrace();
    }
    producer.shutdown();
}

// 消费者
public static void main(String[] args) throws MQClientException {
    DefaultMQPushConsumer consumer = new DefaultMQPushConsumer("rmq-group");
    consumer.setNamesrvAddr("172.16.200.38:9876");
    consumer.setInstanceName("consumer");
    consumer.subscribe("Topic-test", "*");
    consumer.setConsumeFromWhere(ConsumeFromWhere.CONSUME_FROM_FIRST_OFFSET);
    consumer.registerMessageListener(new MessageListenerConcurrently() {
        @Override
        public ConsumeConcurrentlyStatus consumeMessage(List msgs, ConsumeConcurrentlyContext context) {
            for (MessageExt messageExt : msgs) {
                System.out.printf("Receive New Messages: %s At %s %n", new String(messageExt.getBody()), new Date());
            }

            return ConsumeConcurrentlyStatus.CONSUME_SUCCESS;
        }
    });
    //启动消费者
    consumer.start();
}

 控制台打印

// 生产者控制台打印
Wed Jul 26 17:55:45 CST 2023 RocketMQ test msg 0 : executeLocalTransaction:事务执行失败，回滚
Wed Jul 26 17:55:45 CST 2023
7F00000154F073D16E938497DE420000
MessageQueue [topic=Topic-test, brokerName=localhost.localdomain, queueId=1]
SEND_OK
null
93

============================
Wed Jul 26 17:55:46 CST 2023 RocketMQ test msg 1 : executeLocalTransaction:未知状态
Wed Jul 26 17:55:46 CST 2023
7F00000154F073D16E938497E2340001
MessageQueue [topic=Topic-test, brokerName=localhost.localdomain, queueId=2]
SEND_OK
null
94

============================
Wed Jul 26 17:55:47 CST 2023 RocketMQ test msg 2 : executeLocalTransaction:事务执行失败，回滚
Wed Jul 26 17:55:47 CST 2023
7F00000154F073D16E938497E6230002
MessageQueue [topic=Topic-test, brokerName=localhost.localdomain, queueId=3]
SEND_OK
null
95

============================
Wed Jul 26 17:55:48 CST 2023 RocketMQ test msg 3 : executeLocalTransaction:事务执行成功，提交
Wed Jul 26 17:55:48 CST 2023
7F00000154F073D16E938497EA180003
MessageQueue [topic=Topic-test, brokerName=localhost.localdomain, queueId=0]
SEND_OK
null
96

============================
Wed Jul 26 17:55:49 CST 2023 RocketMQ test msg 4 : executeLocalTransaction:事务执行失败，回滚
Wed Jul 26 17:55:49 CST 2023
7F00000154F073D16E938497EE0B0004
MessageQueue [topic=Topic-test, brokerName=localhost.localdomain, queueId=1]
SEND_OK
null
97

============================
Wed Jul 26 17:55:50 CST 2023 RocketMQ test msg 5 : executeLocalTransaction:事务执行成功，提交
Wed Jul 26 17:55:50 CST 2023
7F00000154F073D16E938497F1F70005
MessageQueue [topic=Topic-test, brokerName=localhost.localdomain, queueId=2]
SEND_OK
null
98

============================
Wed Jul 26 17:55:51 CST 2023 RocketMQ test msg 6 : executeLocalTransaction:事务执行成功，提交
Wed Jul 26 17:55:51 CST 2023
7F00000154F073D16E938497F5EC0006
MessageQueue [topic=Topic-test, brokerName=localhost.localdomain, queueId=3]
SEND_OK
null
99

============================
Wed Jul 26 17:55:52 CST 2023 RocketMQ test msg 7 : executeLocalTransaction:事务执行失败，回滚
Wed Jul 26 17:55:52 CST 2023
7F00000154F073D16E938497F9E50007
MessageQueue [topic=Topic-test, brokerName=localhost.localdomain, queueId=0]
SEND_OK
null
100

============================
Wed Jul 26 17:55:53 CST 2023 RocketMQ test msg 8 : executeLocalTransaction:事务执行失败，回滚
Wed Jul 26 17:55:53 CST 2023
7F00000154F073D16E938497FDD30008
MessageQueue [topic=Topic-test, brokerName=localhost.localdomain, queueId=1]
SEND_OK
null
101

============================
Wed Jul 26 17:55:54 CST 2023 RocketMQ test msg 9 : executeLocalTransaction:事务执行成功，提交
Wed Jul 26 17:55:54 CST 2023
7F00000154F073D16E93849801C90009
MessageQueue [topic=Topic-test, brokerName=localhost.localdomain, queueId=2]
SEND_OK
null
102

============================


// 消费者控制台打印
Receive New Messages: Wed Jul 26 17:55:50 CST 2023 RocketMQ test msg 5 At Wed Jul 26 17:55:50 CST 2023 
Receive New Messages: Wed Jul 26 17:55:51 CST 2023 RocketMQ test msg 6 At Wed Jul 26 17:55:51 CST 2023 
Receive New Messages: Wed Jul 26 17:55:54 CST 2023 RocketMQ test msg 9 At Wed Jul 26 17:55:54 CST 2023 

这里消费者只收到了执行事务成功的5,6,9。