Java ScheduledThreadPoolExecutor延迟或周期性执行任务

Java提供的Time类可以周期性地或者延期执行任务，但是有时我们需要并行执行同样的任务，这个时候如果创建多个Time对象会给系统带来负担，解决办法是将定时任务放到线程池中执行。

Java的ScheduledThreadPoolExecutor类实现了ScheduledExecutorService接口中定义的以不同方法执行任务的方法。

之前，我写过一篇关于Java ThreadPoolExecutor的文章中使用了Executors创建线程池。Executors类也提供了工厂方法创建ScheduledThreadPoolExecutor，并且可以设置线程池中的线程。

假设有下面简单的Runnable类

WorkerThread.java:

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

package com.journaldev.threads;   import java.util.Date;   public class WorkerThread implements Runnable{   private String command;        public WorkerThread(String s){          this .command=s;      }        @Override      public void run() {          System.out.println(Thread.currentThread().getName()+ " Start. Time = " + new Date());          processCommand();          System.out.println(Thread.currentThread().getName()+ " End. Time = " + new Date());      }        private void processCommand() {          try {              Thread.sleep( 5000 );          } catch (InterruptedException e) {              e.printStackTrace();          }      }        @Override      public String toString(){          return this .command;      } }

下面的例子中worker线程将被延期10s执行上面的Rnnable类大约花费5s执行任务

ScheduledThreadPool.java:

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

package com.journaldev.threads;   import java.util.Date; import java.util.concurrent.Executors; import java.util.concurrent.ScheduledExecutorService; import java.util.concurrent.TimeUnit;   public class ScheduledThreadPool {        public static void main(String[] args) throws InterruptedException {          ScheduledExecutorService scheduledThreadPool = Executors.newScheduledThreadPool( 5 );            //schedule to run after sometime          System.out.println( "Current Time = " + new Date());          for ( int i= 0 ; i< 3 ; i++){              Thread.sleep( 1000 );              WorkerThread worker = new WorkerThread( "do heavy processing" );              scheduledThreadPool.schedule(worker, 10 , TimeUnit.SECONDS);          }            //add some delay to let some threads spawn by scheduler          Thread.sleep( 30000 );            scheduledThreadPool.shutdown();          while (!scheduledThreadPool.isTerminated()){              //wait for all tasks to finish          }          System.out.println( "Finished all threads" );      }   }

运行上面的程序，可以得到下面的输出，由此可以确认任务在10s后才执行。

1 2 3 4 5 6 7 8

Current Time = Tue Oct 29 15:10:03 IST 2013 pool-1-thread-1 Start. Time = Tue Oct 29 15:10:14 IST 2013 pool-1-thread-2 Start. Time = Tue Oct 29 15:10:15 IST 2013 pool-1-thread-3 Start. Time = Tue Oct 29 15:10:16 IST 2013 pool-1-thread-1 End. Time = Tue Oct 29 15:10:19 IST 2013 pool-1-thread-2 End. Time = Tue Oct 29 15:10:20 IST 2013 pool-1-thread-3 End. Time = Tue Oct 29 15:10:21 IST 2013 Finished all threads

注意到所有的schedule方法都返回了ScheduledFuture实例，可以用于获取线程状态信息和延迟时间。ScheduledFuture接口继承Future接口，更多信息见Java Callable Future Example.

在ScheduledExecutorService中至少有2个方法可用于周期性执行任务。

scheduleAtFixedRate(Runnable command,long initialDelay,long period,TimeUnit unit)

我们可以使用该方法延迟执行任务，设置任务的执行周期。时间周期从线程池中首先开始执行的线程算起，所以假设period为1s，线程执行了5s，那么下一个线程在第一个线程运行完后会很快被执行。

比如下面的代码

1 2 3 4 5 6

for ( int i = 0 ; i < 3 ; i++) {      Thread.sleep( 1000 );      WorkerThread worker = new WorkerThread( "do heavy processing" );      // schedule task to execute at fixed rate      scheduledThreadPool.scheduleAtFixedRate(worker, 0 , 10 ,              TimeUnit.SECONDS);

输出

1 2 3 4 5 6 7 8 9

Current Time = Tue Oct 29 16:10:00 IST 2013 pool-1-thread-1 Start. Time = Tue Oct 29 16:10:01 IST 2013 pool-1-thread-2 Start. Time = Tue Oct 29 16:10:02 IST 2013 pool-1-thread-3 Start. Time = Tue Oct 29 16:10:03 IST 2013 pool-1-thread-1 End. Time = Tue Oct 29 16:10:06 IST 2013 pool-1-thread-2 End. Time = Tue Oct 29 16:10:07 IST 2013 pool-1-thread-3 End. Time = Tue Oct 29 16:10:08 IST 2013 pool-1-thread-1 Start. Time = Tue Oct 29 16:10:11 IST 2013 pool-1-thread-4 Start. Time = Tue Oct 29 16:10:12 IST 2013

scheduleWithFixedDelay(Runnable command,long initialDelay,long delay,TimeUnit unit)

该方法可被用于延迟周期性执行任务，delaytime是线程停止执行到下一次开始执行之间的延迟时间，假设有下面的代码

1 2 3 4 5 6

for ( int i = 0 ; i < 3 ; i++) {      Thread.sleep( 1000 );      WorkerThread worker = new WorkerThread( "do heavy processing" );      scheduledThreadPool.scheduleWithFixedDelay(worker, 0 , 1 ,              TimeUnit.SECONDS); }

输出结果

1 2 3 4 5 6 7 8 9

Current Time = Tue Oct 29 16:14:13 IST 2013 pool-1-thread-1 Start. Time = Tue Oct 29 16:14:14 IST 2013 pool-1-thread-2 Start. Time = Tue Oct 29 16:14:15 IST 2013 pool-1-thread-3 Start. Time = Tue Oct 29 16:14:16 IST 2013 pool-1-thread-1 End. Time = Tue Oct 29 16:14:19 IST 2013 pool-1-thread-2 End. Time = Tue Oct 29 16:14:20 IST 2013 pool-1-thread-1 Start. Time = Tue Oct 29 16:14:20 IST 2013 pool-1-thread-3 End. Time = Tue Oct 29 16:14:21 IST 2013 pool-1-thread-4 Start. Time = Tue Oct 29 16:14:21 IST 2013