异步数据处理

 后台的异步处理

 把需要处理的数据放入阻塞队列中 

public static BlockingQueue<JSONObject> queue = new LinkedBlockingQueue<JSONObject>();

AsyncQueue.queue.put(jsonObject);


然后初始化线程池去执行

public static void init(){
   executor = new ThreadPoolExecutor(100, 200, 30*1000,
         TimeUnit.MILLISECONDS, new LinkedBlockingDeque<Runnable>(5000));
   
   log.info("初始化核心线程数量 "+executor.getCorePoolSize());
   log.info("初始化最大线程数量 "+executor.getMaximumPoolSize());
   
   executor.execute(new SmsAsyncQueueTake(queue));
}


public static void destroyed(){
   executor.shutdown();
}


线程的设计 需要做的事情

public class SmsAsyncQueueTake implements Runnable{
   
   private static Logger log = Logger.getLogger(SmsAsyncQueueTake.class);
   private BlockingQueue<JSONObject> queue;
   
   public SmsAsyncQueueTake(BlockingQueue<JSONObject> queue){
      this.queue=queue;
   }
   
   
   @Override
   public void run() {
      while(!Thread.interrupted()){
         try {
            if(AsyncQueue.executor.getQueue().size()<5000){
               JSONObject sms = queue.take();
               try {
                  AsyncQueue.executor.execute(new SmsAsyncService(sms));
               } catch (RejectedExecutionException e) {
                  log.debug("异步线程执行异常:"+e.getMessage());
               }
               
            }else{
               Thread.sleep(500);
            }
            
         } catch (Exception e) {
            log.error(e.getMessage());
         }
      }
      
   }

}