RocketMQ-负载均衡和广播模式

RocketMQ的消费者消费消息，有2种模式：

• 负载均衡模式
  
• 广播模式
  

首先我们先启动一个生产者：发送了10条消息，主题是TopicTest,tag是TagA

public class SyncProducer {
    public static void main(String[] args) throws MQClientException, UnsupportedEncodingException, RemotingException, InterruptedException, MQBrokerException {
        DefaultMQProducer defaultMQProducer = new DefaultMQProducer();
        //设置生产者组
        defaultMQProducer.setProducerGroup("syncProducer");

        //设置nameserver
        defaultMQProducer.setNamesrvAddr("localhost:9876");

        //启动生产者
        defaultMQProducer.start();

        for (int i = 0; i < 10; i++) {
            //构建消息 topic tag 内容
            Message msg = new Message("TopicTest" /* Topic */,
                    "TagA" /* Tag */,
                    ("Hello RocketMQ " +
                            i).getBytes(RemotingHelper.DEFAULT_CHARSET) /* Message body */
            );
            //同步发送，且返回结果
            SendResult sendResult = defaultMQProducer.send(msg);
            System.out.println("发送结果"+sendResult);
        }
        //关闭生产者
        defaultMQProducer.shutdown();
    }
}
//运行结果
发送结果SendResult [sendStatus=SEND_OK, msgId=7F000001503218B4AAC29874E2040000, offsetMsgId=C0A81FF100002A9F0000000000045402, messageQueue=MessageQueue [topic=TopicTest, brokerName=aarondeMBP, queueId=1], queueOffset=350]
发送结果SendResult [sendStatus=SEND_OK, msgId=7F000001503218B4AAC29874E2230001, offsetMsgId=C0A81FF100002A9F00000000000454CB, messageQueue=MessageQueue [topic=TopicTest, brokerName=aarondeMBP, queueId=2], queueOffset=350]
发送结果SendResult [sendStatus=SEND_OK, msgId=7F000001503218B4AAC29874E2260002, offsetMsgId=C0A81FF100002A9F0000000000045594, messageQueue=MessageQueue [topic=TopicTest, brokerName=aarondeMBP, queueId=3], queueOffset=350]
发送结果SendResult [sendStatus=SEND_OK, msgId=7F000001503218B4AAC29874E22A0003, offsetMsgId=C0A81FF100002A9F000000000004565D, messageQueue=MessageQueue [topic=TopicTest, brokerName=aarondeMBP, queueId=0], queueOffset=350]
发送结果SendResult [sendStatus=SEND_OK, msgId=7F000001503218B4AAC29874E22C0004, offsetMsgId=C0A81FF100002A9F0000000000045726, messageQueue=MessageQueue [topic=TopicTest, brokerName=aarondeMBP, queueId=1], queueOffset=351]
发送结果SendResult [sendStatus=SEND_OK, msgId=7F000001503218B4AAC29874E22E0005, offsetMsgId=C0A81FF100002A9F00000000000457EF, messageQueue=MessageQueue [topic=TopicTest, brokerName=aarondeMBP, queueId=2], queueOffset=351]
发送结果SendResult [sendStatus=SEND_OK, msgId=7F000001503218B4AAC29874E2300006, offsetMsgId=C0A81FF100002A9F00000000000458B8, messageQueue=MessageQueue [topic=TopicTest, brokerName=aarondeMBP, queueId=3], queueOffset=351]
发送结果SendResult [sendStatus=SEND_OK, msgId=7F000001503218B4AAC29874E2320007, offsetMsgId=C0A81FF100002A9F0000000000045981, messageQueue=MessageQueue [topic=TopicTest, brokerName=aarondeMBP, queueId=0], queueOffset=351]
发送结果SendResult [sendStatus=SEND_OK, msgId=7F000001503218B4AAC29874E2340008, offsetMsgId=C0A81FF100002A9F0000000000045A4A, messageQueue=MessageQueue [topic=TopicTest, brokerName=aarondeMBP, queueId=1], queueOffset=352]
发送结果SendResult [sendStatus=SEND_OK, msgId=7F000001503218B4AAC29874E2360009, offsetMsgId=C0A81FF100002A9F0000000000045B13, messageQueue=MessageQueue [topic=TopicTest, brokerName=aarondeMBP, queueId=2], queueOffset=352]
14:29:56.930 [NettyClientSelector_1] INFO RocketmqRemoting - closeChannel: close the connection to remote address[127.0.0.1:9876] result: true
14:29:56.932 [NettyClientSelector_1] INFO RocketmqRemoting - closeChannel: close the connection to remote address[127.0.0.1:9876] result: true
14:29:56.932 [NettyClientSelector_1] INFO RocketmqRemoting - closeChannel: close the connection to remote address[192.168.31.241:10911] result: true

接下来，我们启动2个消费者，分别用负载均衡模式和广播模式去进行消费信息：

负载均衡(或叫集群模式)

public class ClusterConsumer {
    public static void main(String[] args) throws MQClientException {
        DefaultMQPushConsumer consumer = new DefaultMQPushConsumer("clusterConsumer");

        consumer.setNamesrvAddr("localhost:9876");

        //设置集群模式，也就是负载均衡模式
        consumer.setMessageModel(MessageModel.CLUSTERING);
        //订阅主题和标签
        consumer.subscribe("TopicTest","TagA");


        consumer.registerMessageListener(new MessageListenerConcurrently() {
            @Override
            public ConsumeConcurrentlyStatus consumeMessage(List<MessageExt> msgs, ConsumeConcurrentlyContext context) {
                for (MessageExt msg : msgs) {
                    System.out.println("消费信息:"+new String(msg.getBody()));
                }
                return ConsumeConcurrentlyStatus.CONSUME_SUCCESS;
            }
        });

        consumer.start();
    }
}
//运行2个实例，结果如下：
实例1：
消费信息:Hello RocketMQ 3
消费信息:Hello RocketMQ 2
消费信息:Hello RocketMQ 7
消费信息:Hello RocketMQ 6

实例2：
消费信息:Hello RocketMQ 1
消费信息:Hello RocketMQ 0
消费信息:Hello RocketMQ 4
消费信息:Hello RocketMQ 5
消费信息:Hello RocketMQ 8
消费信息:Hello RocketMQ 9

广播模式

public class BoardConsumer {
    public static void main(String[] args) throws MQClientException {
        DefaultMQPushConsumer consumer = new DefaultMQPushConsumer("clusterConsumer");

        consumer.setNamesrvAddr("localhost:9876");

        //设置集群模式，也就是负载均衡模式
        consumer.setMessageModel(MessageModel.BROADCASTING);
        //订阅主题和标签
        consumer.subscribe("TopicTest","TagA");


        consumer.registerMessageListener(new MessageListenerConcurrently() {
            @Override
            public ConsumeConcurrentlyStatus consumeMessage(List<MessageExt> msgs, ConsumeConcurrentlyContext context) {
                for (MessageExt msg : msgs) {
                    System.out.println("消费信息:"+new String(msg.getBody()));
                }
                return ConsumeConcurrentlyStatus.CONSUME_SUCCESS;
            }
        });

        consumer.start();
    }
//运行2个实例，结果如下：
实例1：
消费信息:Hello RocketMQ 0
消费信息:Hello RocketMQ 1
消费信息:Hello RocketMQ 2
消费信息:Hello RocketMQ 3
消费信息:Hello RocketMQ 5
消费信息:Hello RocketMQ 9
消费信息:Hello RocketMQ 4
消费信息:Hello RocketMQ 6
消费信息:Hello RocketMQ 7
消费信息:Hello RocketMQ 8


实例2：
消费信息:Hello RocketMQ 2
消费信息:Hello RocketMQ 3
消费信息:Hello RocketMQ 1
消费信息:Hello RocketMQ 0
消费信息:Hello RocketMQ 5
消费信息:Hello RocketMQ 9
消费信息:Hello RocketMQ 4
消费信息:Hello RocketMQ 8
消费信息:Hello RocketMQ 7
消费信息:Hello RocketMQ 6

今天分享了消费者负载均衡模式和广播模式。
在生产中，一般都是用负载均衡模式。广播模式比较少用。但还是得具体场景具体分析。

后续文章

• RocketMQ-入门(已更新)
• RocketMQ-消息发送(已更新)
• RocketMQ-消费信息
• RocketMQ-消费者的广播模式和集群模式（已更新）
• RocketMQ-顺序消息
• RocketMQ-延迟消息
• RocketMQ-批量消息
• RocketMQ-过滤消息
• RocketMQ-事务消息
• RocketMQ-消息存储
• RocketMQ-高可用
• RocketMQ-高性能
• RocketMQ-主从复制
• RocketMQ-刷盘机制
• RocketMQ-幂等性
• RocketMQ-消息重试
• RocketMQ-死信队列
  ...

欢迎各位入(guan)股(zhu)，后续文章干货多多。