SpringCloud+RocketMQ实现分布式事务的实践

start mqnameserv.cmd

start mqbroker.cmd -n 127.0.0.1:9876 autoCreateTopicEnable=true

DROP TABLE IF EXISTS `spring_cloud_config`;

CREATE TABLE `spring_cloud_config` (
  `id` int(11) unsigned NOT NULL AUTO_INCREMENT,
  `akey` varchar(30) DEFAULT NULL,
  `avalue` varchar(128) DEFAULT NULL,
  `application` varchar(30) DEFAULT NULL,
  `aprofile` varchar(30) DEFAULT NULL,
  `label` varchar(30) DEFAULT NULL,
  PRIMARY KEY (`id`)
) ENGINE=InnoDB DEFAULT CHARSET=latin1;



INSERT INTO `spring_cloud_config` (`id`, `akey`, `avalue`, `application`, `aprofile`, `label`)
VALUES
    (2,'name_server','127.0.0.1:9876','product-service','dev','dev'),
    (3,'name_server','127.0.0.1:9876','order-service','dev','dev'),
    (4,'order_topic','order_topic','order-service','dev','dev'),
    (5,'order_topic','order_topic','product-service','dev','dev');

package com.hqs.chargeorderservice.mqservice;

import com.alibaba.fastjson.JSONObject;
import com.hqs.chargeorderservice.config.Jms;
import com.hqs.chargeorderservice.service.ProduceOrderService;
import lombok.extern.slf4j.Slf4j;
import org.apache.rocketmq.client.exception.MQClientException;
import org.apache.rocketmq.client.producer.LocalTransactionState;
import org.apache.rocketmq.client.producer.TransactionListener;
import org.apache.rocketmq.client.producer.TransactionMQProducer;
import org.apache.rocketmq.common.message.Message;
import org.apache.rocketmq.common.message.MessageExt;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.stereotype.Component;

import java.util.concurrent.*;


/**
 *
 * @Description: 分布式事务RocketMQ 生产者
 */
@Slf4j
@Component
public class TransactionProducer {

    /**
     * 需要自定义事务监听器 用于 事务的二次确认 和 事务回查
     */
    private TransactionListener transactionListener ;

    /**
     * 这里的生产者和之前的不一样
     */
    private TransactionMQProducer producer = null;

    /**
     * 官方建议自定义线程 给线程取自定义名称 发现问题更好排查
     */
    private ExecutorService executorService = new ThreadPoolExecutor(2, 5, 100, TimeUnit.SECONDS,
            new ArrayBlockingQueue<Runnable>(2000), new ThreadFactory() {
        @Override
        public Thread newThread(Runnable r) {
            Thread thread = new Thread(r);
            thread.setName("client-transaction-msg-check-thread");
            return thread;
        }

    });

    public TransactionProducer(@Autowired Jms jms, @Autowired ProduceOrderService produceOrderService) {
        transactionListener = new TransactionListenerImpl(produceOrderService);
        // 初始化 事务生产者
        producer = new TransactionMQProducer(jms.getOrderTopic());
        // 添加服务器地址
        producer.setNamesrvAddr(jms.getNameServer());
        // 添加事务监听器
        producer.setTransactionListener(transactionListener);
        // 添加自定义线程池
        producer.setExecutorService(executorService);

        start();
    }

    public TransactionMQProducer getProducer() {
        return this.producer;
    }

    /**
     * 对象在使用之前必须要调用一次，只能初始化一次
     */
    public void start() {
        try {
            this.producer.start();
        } catch (MQClientException e) {
            e.printStackTrace();
        }
    }

    /**
     * 一般在应用上下文，使用上下文监听器，进行关闭
     */
    public void shutdown() {
        this.producer.shutdown();
    }
}

/**
 * @Description: 自定义事务监听器
 */
@Slf4j
class TransactionListenerImpl implements TransactionListener {

    @Autowired
    private ProduceOrderService produceOrderService ;

    public TransactionListenerImpl( ProduceOrderService produceOrderService) {
        this.produceOrderService = produceOrderService;
    }

    @Override
    public LocalTransactionState executeLocalTransaction(Message msg, Object arg) {
        log.info("=========本地事务开始执行=============");
        String message = new String(msg.getBody());
        JSONObject jsonObject = JSONObject.parseObject(message);
        Integer productId = jsonObject.getInteger("productId");
        Integer total = jsonObject.getInteger("total");
        Integer tradeNo = jsonObject.getInteger("tradeNo");
        int userId = Integer.parseInt(arg.toString());
        //模拟执行本地事务begin=======
        /**
         * 本地事务执行会有三种可能
         * 1、commit 成功
         * 2、Rollback 失败
         * 3、网络等原因服务宕机收不到返回结果
         */
        log.info("本地事务执行参数,用户id={},商品ID={},销售库存={}",userId,productId,total);
        int result = produceOrderService.save(userId, productId, total, tradeNo);
        //模拟执行本地事务end========
        //TODO 实际开发下面不需要我们手动返回，而是根据本地事务执行结果自动返回
        //1、二次确认消息，然后消费者可以消费
        if (result == 0) {
            return LocalTransactionState.COMMIT_MESSAGE;
        }
        //2、回滚消息，Broker端会删除半消息
        if (result == 1) {
            return LocalTransactionState.ROLLBACK_MESSAGE;
        }
        //3、Broker端会进行回查消息
        if (result == 2) {
            return LocalTransactionState.UNKNOW;
        }
        return LocalTransactionState.COMMIT_MESSAGE;
    }

    /**
     * 只有上面接口返回 LocalTransactionState.UNKNOW 才会调用查接口被调用
     *
     * @param msg 消息
     * @return
     */
    @Override
    public LocalTransactionState checkLocalTransaction(MessageExt msg) {
        log.info("==========回查接口=========");
        String key = msg.getKeys();
        //TODO 1、必须根据key先去检查本地事务消息是否完成。
        /**
         * 因为有种情况就是：上面本地事务执行成功了，但是return LocalTransactionState.COMMIT_MESSAG的时候
         * 服务挂了，那么最终 Brock还未收到消息的二次确定，还是个半消息 ，所以当重新启动的时候还是回调这个回调接口。
         * 如果不先查询上面本地事务的执行情况 直接在执行本地事务，那么就相当于成功执行了两次本地事务了。
         */
        // TODO 2、这里返回要么commit 要么rollback。没有必要在返回 UNKNOW
        return LocalTransactionState.COMMIT_MESSAGE;
    }
}