缓存型线程池 newCachedThreadPool

旧城等待， 2024-03-22 19:34 81阅读 0赞

### newCachedThreadPool ###

public static ExecutorService newCachedThreadPool() {
        
        return new ThreadPoolExecutor(0, 
        							  Integer.MAX_VALUE,
                                      60L, TimeUnit.SECONDS,
                                      new SynchronousQueue<Runnable>());
    }

<table> 
 <thead> 
  <tr> 
   <th align="left">CachedThreadPool是一个会根据需要创建新线程的线程池</th> 
  </tr> 
 </thead> 
 <tbody> 
  <tr> 
   <td align="left">CachedThreadPool的corePoolSize被设置为<code>0</code>，即corePool为空；</td> 
  </tr> 
  <tr> 
   <td align="left">maximumPoolSize被设置为<code>Integer.MAX_VALUE</code>，即maximumPool是无界的。</td> 
  </tr> 
  <tr> 
   <td align="left">这里把keepAliveTime设置为<code>60L</code>，意味着CachedThreadPool中的空闲线程等待新任务的最长时间为60秒，空闲线程超过60秒后将会被终止。</td> 
  </tr> 
  <tr> 
   <td align="left">FixedThreadPool和SingleThreadExecutor使用无界队列<code>LinkedBlockingQueue</code>作为线程池的工作队列。CachedThreadPool使用没有容量的<code>SynchronousQueue</code>作为线程池的工作队列，但CachedThreadPool的maximumPool是无界的。这意味着，<strong>如果主线程提交任务的速度高于maximumPool中线程处理任务的速度时，CachedThreadPool会不断创建新线程。极端情况下，CachedThreadPool会因为创建过多线程而耗尽CPU和内存资源</strong>。</td> 
  </tr> 
  <tr> 
   <td align="left"></td> 
  </tr> 
  <tr> 
   <td align="left">CachedThreadPool的execute()方法的执行示意图如下图所示</td> 
  </tr> 
 </tbody> 
</table>

![在这里插入图片描述][7125d4d997c24e9b9fe948eb0c19ce7e.png_pic_center]

<table> 
 <thead> 
  <tr> 
   <th align="left"></th> 
  </tr> 
 </thead> 
 <tbody> 
  <tr> 
   <td align="left">1）首先执行<code>SynchronousQueue.offer（Runnable task）</code>。如果当前maximumPool中有空闲线程正在执行<code>SynchronousQueue.poll（keepAliveTime，TimeUnit.NANOSECONDS）</code>，那么主线程执行offer操作与空闲线程执行的poll操作<code>配对成功</code>，主线程把任务交给空闲线程执行，execute()方法执行完成；否则执行下面的步骤2）。</td> 
  </tr> 
  <tr> 
   <td align="left">2）当初始maximumPool<code>为空</code>，或者maximumPool中当前<code>没有空闲线程</code>时，将没有线程执行SynchronousQueue.poll（keepAliveTime，TimeUnit.NANOSECONDS）。这种情况下，步骤1）将失败。此时CachedThreadPool会<code>创建一个新线程执行任务</code>，execute()方法执行完成。</td> 
  </tr> 
  <tr> 
   <td align="left">3）在步骤2）中新创建的线程将任务执行完后，会执行SynchronousQueue.poll（keepAliveTime，TimeUnit.NANOSECONDS）。**这个poll操作会让空闲线程最多在SynchronousQueue中等待60秒钟。如果60秒钟内主线程提交了一个新任务（主线程执行步骤1）），那么这个空闲线程将执行主线程提交的新任务；否则，这个空闲线程将终止。**由于空闲60秒的空闲线程会被终止，因此长时间保持空闲的CachedThreadPool不会使用任何资源。</td> 
  </tr> 
 </tbody> 
</table>

####  CachedThreadPool的任务传递示意图  ####

![在这里插入图片描述][598fb7efdfca45aa9d4bc029c8241bc0.png_pic_center]

<table> 
 <thead> 
  <tr> 
   <th align="left">SynchronousQueue是一个没有容量的阻塞队列。<code>每个插入操作必须等待另一个线程的对应移除操作，反之亦然</code>。CachedThreadPool使用SynchronousQueue，把主线程提交的任务传递给空闲线程执行。</th> 
  </tr> 
 </thead> 
</table>

\-----------------------------------------------------------------------------读书笔记摘自 书名：Java并发编程的艺术 作者：方腾飞；魏鹏；程晓明

### 常见用途 ###

参考：[Executor框架的成员][Executor]

--------------------

### 线程任务 ###

public class Task implements Runnable {
        
        private final String name;
    
        public Task(String name) {
        
            this.name = name;
        }
    
        @Override
        public void run() {
        
            System.out.println(Thread.currentThread().getName() + " → " + name + " Start Time = " + new Date());
            System.out.println(Thread.currentThread().getName() + " → " + name + " End   Time = " + new Date());
        }
    }

### 缓存型线程池 ###

public class cachedThreadPool {
        
        public static void main(String[] args) throws InterruptedException, ExecutionException {
        
    
            System.out.println(Thread.currentThread().getName() + "线程: Start at: " + new Date());
    
            ExecutorService exec = Executors.newCachedThreadPool();
    
            for (int i = 1; i < 10; i++) {
        
                System.out.println("添加了第" + i + "个任务类");
                Thread.sleep(2);
                exec.execute(new Task("线程名字" + i));
            }
    
            exec.shutdown();
    
            System.out.println(Thread.currentThread().getName() + " 线程: Finished all threads at：" + new Date());
        }
    }

--------------------

### 执行结果分析 ###

<table> 
 <thead> 
  <tr> 
   <th align="left">执行结果</th> 
  </tr> 
 </thead> 
 <tbody> 
  <tr> 
   <td align="left">main线程: Start at: Fri May 19 20:55:45 CST 2023</td> 
  </tr> 
  <tr> 
   <td align="left">添加了第1个任务类</td> 
  </tr> 
  <tr> 
   <td align="left">添加了第2个任务类</td> 
  </tr> 
  <tr> 
   <td align="left">添加了第3个任务类</td> 
  </tr> 
  <tr> 
   <td align="left">添加了第4个任务类</td> 
  </tr> 
  <tr> 
   <td align="left">添加了第5个任务类</td> 
  </tr> 
  <tr> 
   <td align="left">pool-1-thread-1 → 线程名字1 Start Time = Fri May 19 20:55:45 CST 2023</td> 
  </tr> 
  <tr> 
   <td align="left">pool-1-thread-4 → 线程名字4 Start Time = Fri May 19 20:55:45 CST 2023</td> 
  </tr> 
  <tr> 
   <td align="left">pool-1-thread-2 → 线程名字2 Start Time = Fri May 19 20:55:45 CST 2023</td> 
  </tr> 
  <tr> 
   <td align="left">pool-1-thread-3 → 线程名字3 Start Time = Fri May 19 20:55:45 CST 2023</td> 
  </tr> 
  <tr> 
   <td align="left">pool-1-thread-1 → 线程名字1 End Time = Fri May 19 20:55:45 CST 2023</td> 
  </tr> 
  <tr> 
   <td align="left">pool-1-thread-4 → 线程名字4 End Time = Fri May 19 20:55:45 CST 2023</td> 
  </tr> 
  <tr> 
   <td align="left">pool-1-thread-3 → 线程名字3 End Time = Fri May 19 20:55:45 CST 2023</td> 
  </tr> 
  <tr> 
   <td align="left">pool-1-thread-2 → 线程名字2 End Time = Fri May 19 20:55:45 CST 2023</td> 
  </tr> 
  <tr> 
   <td align="left">添加了第6个任务类</td> 
  </tr> 
  <tr> 
   <td align="left">pool-1-thread-3 → 线程名字5 Start Time = Fri May 19 20:55:45 CST 2023</td> 
  </tr> 
  <tr> 
   <td align="left">pool-1-thread-3 → 线程名字5 End Time = Fri May 19 20:55:45 CST 2023</td> 
  </tr> 
  <tr> 
   <td align="left">添加了第7个任务类</td> 
  </tr> 
  <tr> 
   <td align="left">pool-1-thread-3 → 线程名字6 Start Time = Fri May 19 20:55:45 CST 2023</td> 
  </tr> 
  <tr> 
   <td align="left">pool-1-thread-3 → 线程名字6 End Time = Fri May 19 20:55:45 CST 2023</td> 
  </tr> 
  <tr> 
   <td align="left">添加了第8个任务类</td> 
  </tr> 
  <tr> 
   <td align="left">pool-1-thread-3 → 线程名字7 Start Time = Fri May 19 20:55:45 CST 2023</td> 
  </tr> 
  <tr> 
   <td align="left">pool-1-thread-3 → 线程名字7 End Time = Fri May 19 20:55:45 CST 2023</td> 
  </tr> 
  <tr> 
   <td align="left">添加了第9个任务类</td> 
  </tr> 
  <tr> 
   <td align="left">pool-1-thread-3 → 线程名字8 Start Time = Fri May 19 20:55:45 CST 2023</td> 
  </tr> 
  <tr> 
   <td align="left">pool-1-thread-3 → 线程名字8 End Time = Fri May 19 20:55:45 CST 2023</td> 
  </tr> 
  <tr> 
   <td align="left">pool-1-thread-3 → 线程名字9 Start Time = Fri May 19 20:55:45 CST 2023</td> 
  </tr> 
  <tr> 
   <td align="left">pool-1-thread-3 → 线程名字9 End Time = Fri May 19 20:55:45 CST 2023</td> 
  </tr> 
  <tr> 
   <td align="left">main 线程: Finished all threads at：Fri May 19 20:55:45 CST 2023</td> 
  </tr> 
  <tr> 
   <td align="left"></td> 
  </tr> 
  <tr> 
   <td align="left"><strong>从控制台结果可以看出：</strong></td> 
  </tr> 
  <tr> 
   <td align="left">发现<strong>reuse</strong>的线程，线程池创建<code>4</code>个线程来 执行这<code>9</code>个任务</td> 
  </tr> 
  <tr> 
   <td align="left">主线程的执行与线程池里的线程分开，有可能主线程结束了，但是线程池还在运行</td> 
  </tr> 
  <tr> 
   <td align="left">放入线程池的线程并不一定会按其放入的先后而顺序执行</td> 
  </tr> 
 </tbody> 
</table>

[7125d4d997c24e9b9fe948eb0c19ce7e.png_pic_center]: https://image.dandelioncloud.cn/pgy_files/images/2024/03/18/ba81e34b10644d099cf04f9a61663a70.png
[598fb7efdfca45aa9d4bc029c8241bc0.png_pic_center]: https://image.dandelioncloud.cn/pgy_files/images/2024/03/18/48f22be9cd8948ffb7268732623c5d35.png
[Executor]: https://blog.csdn.net/weixin_37646636/article/details/131028935