DelayQueue使用

DelayQueue使用:
     DelayQueue是BlockingQueue的一种，所以它是线程安全的。
     时间到期后元素才能被取出使用,所以可以通过设置时间来控制元素输出顺序。
     使用DelayQueue首先需要实现Delayed：
     eg:
     @Data
     public class DelayItem  implements Delayed{

         /* 触发时间*/
         private long time;

         String delayKey;

         public DelayItem(String delayKey, long time, TimeUnit unit) {
             this.delayKey = delayKey;
             this.time = System.currentTimeMillis() + (time > 0 ? unit.toMillis(time) : 0);
         }

         @Override
         public long getDelay(TimeUnit unit) {
             return time - System.currentTimeMillis();
         }

         @Override
         public int compareTo(Delayed o) {
             DelayItem item = (DelayItem) o;
             long diff = this.time - item.time;
             if (diff <= 0) {// 改成>=会造成问题
                 return -1;
             } else {
                 return 1;
             }
         }
     }

    public class MemoryData {
        public static DelayQueue<DelayItem> delayQueue = new DelayQueue<>();
    }

     生产者可以通过redis记录详细数据,delatItem只记录redis的key。
     eg:
     redisBaseService.sAdd(key, value);
     DelayItem delayItem = new DelayItem(key, 1, TimeUnit.MINUTES);
     MemoryData.delayQueue.put(delayItem);


     消费者可以增加监听，每分钟监听延迟队列里是否存在到期数据，甚至启停服务时，将未处理的数据放入redis，下次起
     服务时,从redis中获取对应数据。
      eg:
      启动类:
      ConfigurableApplicationContext run =
                         SpringApplication.run(Application.class, args);
      run.addApplicationListener(new DataMergeContainer());


      使用类:
      @Component
      public class DataMergeContainer implements ApplicationListener<ApplicationStartedEvent>, DisposableBean {

          private static final Logger log = LoggerFactory.getLogger(DataMergeContainer.class);

          @Autowired
          private RedisBaseService redisBaseService;

          private void peakData() {
              while (MemoryData.delayQueue.isEmpty()) {
                  try {
                      Thread.sleep(60000);
                  } catch (InterruptedException e) {
                      log.error("peakData sleep error", e);
                  }
              }


              if (MemoryData.delayQueue.size() > 0) {
                  for (int i = 0; i < MemoryData.delayQueue.size(); i++) {
                      DelayItem take;
                      try {
                          take = MemoryData.delayQueue.take();
                          String delayKey = take.getDelayKey();
                          if (!StringUtils.isBlank(delayKey)) {
                             //业务使用
                          }

                      } catch (InterruptedException e) {
                          log.info("peakData getData error", e);
                      }
                  }
              }
              //递归处置
              peakData();
          }


          @Override
          public void onApplicationEvent(ApplicationStartedEvent event) {
              log.info("DataMergeContainer start.....");
              while (redisBaseService.sSize("上次关闭服务记录数据的key") > 0) {
                  String key= (String) redisBaseService.sPop("上次关闭服务记录数据的key");
                  DelayItem delayItem = new DelayItem(key, 1, TimeUnit.MINUTES);
                  MemoryData.delayQueue.put(delayItem);
              }
              new Thread(() -> {
                  peakData();
              }).start();
              log.info("peakData getData start...");
          }

          @Override
          public void destroy() {
                for (int i = 0; i < MemoryData.delayQueue.size(); i++) {
                  DelayItem take;
                  try {
                      take = MemoryData.delayQueue.take();
                      if (StringUtils.isNotBlank(take.getDelayKey())) {
                          redisBaseService.sAdd("未处理完的服务记录数据的key", take.getDelayKey());
                      }
                  } catch (InterruptedException e) {
                      log.info("peakData getData error", e);
                  }
              }
          }
      }

由于项目使用的是rabbitmq而非RocketMQ,否则使用RocketMQ的延迟消息机制会更方便,根据业务场景不同，可以使用不同的实现机制
包括不限于延迟队列，消息，定时任务等