ActiveMQ 与 Spring Boot 整合实践指南

消息队列在现代应用中扮演着重要角色，尤其是在处理高并发场景时，能够有效地解耦系统，提升整体性能。ActiveMQ 作为一个成熟的开源消息队列系统，常与 Spring Boot 应用进行集成。本文将从安装部署、开发实践到实际应用场景，详细介绍 ActiveMQ 与 Spring Boot 的整合方法。

1. ActiveMQ 安装与启动

1.1 安装 ActiveMQ

ln -s /opt/activemq/bin/activemq /etc/init.d/activemqchkconfig --add activemq

1.2 启动服务

service activemq start

1.3 关闭服务

service activemq stop

1.4 通过 netstat 查看端口

netstat -tuln | grep 61616

ActiveMQ 提供两个重要端口：

消息队列服务端口：61616

控制台管理端口：8161

2. 在 Spring Boot 项目中整合 ActiveMQ

2.1 导入依赖


       
    
     org.springframework.boot
        
    
     spring-boot-starter-activemq
        
            
                 
      
       org.slf4j
                  
      
       slf4j-log4j12
              
         
    
   
       
    
     org.apache.activemq
        
    
     activemq-pool
        
    
     5.15.2
        
            
                 
      
       org.slf4j
                  
      
       slf4j-log4j12

2.2 依赖配置

@Configurationpublic class ActiveMQConfig {    @Value("${spring.activemq.broker-url:disabled}")    private String brokerURL;    @Value("${activemq.listener.enable:disabled}")    private String listenerEnable;    @Bean    public ActiveMQUtil getActiveMQUtil() throws JMSException {        if (brokerURL.equals("disabled")) {            return null;        }        ActiveMQUtil activeMQUtil = new ActiveMQUtil();        activeMQUtil.init(brokerURL);        return activeMQUtil;    }    @Bean(name = "jmsQueueListener")    public DefaultJmsListenerContainerFactory jmsQueueListenerContainerFactory(            ActiveMQConnectionFactory activeMQConnectionFactory) {        DefaultJmsListenerContainerFactory factory = new DefaultJmsListenerContainerFactory();        if (!listenerEnable.equals("true")) {            return null;        }        factory.setConnectionFactory(activeMQConnectionFactory);        factory.setConcurrency("5");        factory.setRecoveryInterval(5000L);        factory.setSessionTransacted(false);        factory.setSessionAcknowledgeMode(Session.CLIENT_ACKNOWLEDGE);        return factory;    }    @Bean    public ActiveMQConnectionFactory activeMQConnectionFactory() {        return new ActiveMQConnectionFactory(brokerURL);    }}

2.3 工具类实现

public class ActiveMQUtil {    private PooledConnectionFactory pooledConnectionFactory;    public ConnectionFactory init(String brokerUrl) {        ActiveMQConnectionFactory factory = new ActiveMQConnectionFactory(brokerUrl);        pooledConnectionFactory = new PooledConnectionFactory(factory);        pooledConnectionFactory.setReconnectOnException(true);        pooledConnectionFactory.setMaxConnections(5);        pooledConnectionFactory.setExpiryTimeout(10000);        return pooledConnectionFactory;    }    public ConnectionFactory getConnectionFactory() {        return pooledConnectionFactory;    }}

3. 消息生产与消费示例

3.1 产生消息

public static void main(String[] args) {    ConnectionFactory connect = new ActiveMQConnectionFactory("tcp://192.168.67.163:61616");    try {        Connection connection = connect.createConnection();        connection.start();        Session session = connection.createSession(true, Session.SESSION_TRANSACTED);        Queue testqueue = session.createQueue("TEST1");        MessageProducer producer = session.createProducer(testqueue);        TextMessage textMessage = new ActiveMQTextMessage();        textMessage.setText("今天天气真好！");        producer.setDeliveryMode(DeliveryMode.PERSISTENT);        producer.send(textMessage);        session.commit();        connection.close();    } catch (JMSException e) {        e.printStackTrace();    }}

3.2 消费消息

public static void main(String[] args) {    ConnectionFactory connect = new ActiveMQConnectionFactory(            ActiveMQConnection.DEFAULT_USER,            ActiveMQConnection.DEFAULT_PASSWORD,            "tcp://192.168.67.163:61616");    try {        Connection connection = connect.createConnection();        connection.start();        Session session = connection.createSession(false, Session.AUTO_ACKNOWLEDGE);        Destination testqueue = session.createQueue("TEST1");        MessageConsumer consumer = session.createConsumer(testqueue);        consumer.setMessageListener(new MessageListener() {            @Override            public void onMessage(Message message) {                if (message instanceof TextMessage) {                    try {                        String text = ((TextMessage) message).getText();                        System.out.println(text);                        // session.rollback();                    } catch (JMSException e) {                        e.printStackTrace();                    }                }            }        });    } catch (Exception e) {        e.printStackTrace();    }}

4. 事务控制与持久化

4.1 事务控制

通过 Session 的事务管理实现消息生产和消费的原子性。

4.2 持久化设置

producer.setDeliveryMode(DeliveryMode.PERSISTENT);

持久化消息在 ActiveMQ 宕机时不会丢失，但会增加系统资源消耗。

5. 在支付业务模块中的应用

5.1 系统设计

支付模块通过消息队列通知订单系统，实现异步状态更新。

5.2 配置

在 application.properties 中配置：

spring.activemq.broker-url=tcp://mq.server.com:61616activemq.listener.enable=true

5.3 代码实现

public void sendPaymentResult(String orderId, String result) {    ConnectionFactory connectionFactory = activeMQUtil.getConnectionFactory();    Connection connection = null;    try {        connection = connectionFactory.createConnection();        connection.start();        Session session = connection.createSession(true, Session.SESSION_TRANSACTED);        Queue paymentResultQueue = session.createQueue("PAYMENT_RESULT_QUEUE");        MapMessage mapMessage = new ActiveMQMapMessage();        mapMessage.setString("orderId", orderId);        mapMessage.setString("result", result);        MessageProducer producer = session.createProducer(paymentResultQueue);        producer.send(mapMessage);        session.commit();        producer.close();        connection.close();    } catch (JMSException e) {        e.printStackTrace();    }}

6. 订单模块的消息消费

6.1 配置

spring.activemq.broker-url=tcp://mq.server.com:61616activemq.listener.enable=true

6.2 消息消费逻辑

@JmsListener(destination = "PAYMENT_RESULT_QUEUE", containerFactory = "jmsQueueListener")public void consumePaymentResult(MapMessage mapMessage) throws JMSException {    String orderId = mapMessage.getString("orderId");    String result = mapMessage.getString("result");    if (!"success".equals(result)) {        orderService.updateProcessStatus(orderId, ProcessStatus.PAY_FAIL);    } else {        orderService.updateProcessStatus(orderId, ProcessStatus.PAID);    }    orderService.sendOrderResult(orderId);}

7. 减库存通知

7.1 消息发送逻辑

@Transactionalpublic void sendOrderResult(String orderId) {    OrderInfo orderInfo = getOrderInfo(orderId);    Map
   
     messageMap = initWareOrderMessage(orderInfo);    String wareOrderJson = JSON.toJSONString(messageMap);    Session session = null;    try {        Connection conn = activeMQUtil.getConnection();        session = conn.createSession(true, Session.SESSION_TRANSACTED);        Queue queue = session.createQueue("ORDER_RESULT_QUEUE");        MessageProducer producer = session.createProducer(queue);        TextMessage message = new ActiveMQTextMessage();        message.setText(wareOrderJson);        producer.send(message);        updateProcessStatus(orderInfo.getId(), ProcessStatus.NOTIFIED_WARE);        session.commit();        producer.close();        conn.close();    } catch (JMSException e) {        e.printStackTrace();    }}

7.2 消息数据组织

public Map
   
     initWareOrderMessage(OrderInfo orderInfo) {    HashMap
    
      hashMap = new HashMap<>();    hashMap.put("orderId", orderInfo.getId());    hashMap.put("consignee", orderInfo.getConsignee());    hashMap.put("consigneeTel", orderInfo.getConsigneeTel());    hashMap.put("orderComment", orderInfo.getOrderComment());    hashMap.put("orderBody", orderInfo.getOrderSubject());    hashMap.put("deliveryAddress", orderInfo.getDeliveryAddress());    hashMap.put("paymentWay", "2"); // 1 货到付款 2 在线支付    hashMap.put("wareId", orderInfo.getWareId());        List
     
      
       > details = new ArrayList<>();    List
       
         orderDetailList = orderInfo.getOrderDetailList();    for (OrderDetail orderDetail : orderDetailList) {        HashMap
        
          detailMap = new HashMap<>(); detailMap.put("skuId", orderDetail.getSkuId()); detailMap.put("skuNum", "" + orderDetail.getSkuNum()); detailMap.put("skuName", orderDetail.getSkuName()); details.add(detailMap); } hashMap.put("details", details); return hashMap;}

通过以上实践，可以看到消息队列在业务逻辑中的重要作用，能够有效地解耦系统组件，提升系统的可扩展性和可维护性。

转载地址：http://aagfk.baihongyu.com/

你可能感兴趣的文章

multivariate_normal TypeError: ufunc ‘add‘ output (typecode ‘O‘) could not be coerced to provided……