Java-五种线程池，四种拒绝策略，三种阻塞队列

三种阻塞队列

BlockingQueue<Runnable> workQueue = null;
workQueue = new ArrayBlockingQueue<>(5);//基于数组的先进先出队列，有界
workQueue = new LinkedBlockingQueue<>();//基于链表的先进先出队列，无界
workQueue = new SynchronousQueue<>();//无缓冲的等待队列，无界

四种拒绝策略

RejectedExecutionHandler rejected = null;
rejected = new ThreadPoolExecutor.AbortPolicy();//默认，队列满了丢任务抛出异常
rejected = new ThreadPoolExecutor.DiscardPolicy();//队列满了丢任务不异常
rejected = new ThreadPoolExecutor.DiscardOldestPolicy();//将最早进入队列的任务删，之后再尝试加入队列
rejected = new ThreadPoolExecutor.CallerRunsPolicy();//如果添加到线程池失败，那么主线程会自己去执行该任务

如何创建线程池？

直接使用ThreadPoolExecutor类的构造方法就可以创建线程池。我们来看下构造参数：

/**
*  corePoolSize    核心线程数
*  maximumPoolSize 最大线程数
*  keepAliveTime   idle线程存活时间
*  unit            上个参数的单位
*  workQueue       线程对象的缓冲队列
*  threadFactory   生成线程的工厂(可选
*  handler         达到容量后的回调(可选)
*/
public ThreadPoolExecutor(int corePoolSize,
                              int maximumPoolSize,
                              long keepAliveTime,
                              TimeUnit unit,
                              BlockingQueue<Runnable> workQueue,
                              ThreadFactory threadFactory,
                              RejectedExecutionHandler handler)

五种线程池

ExecutorService threadPool = null;

threadPool = Executors.newSingleThreadExecutor();//单线程的线程池，只有一个线程在工作，阻塞队列使用的是LinkedBlockingQueue

threadPool = Executors.newFixedThreadPool(3);//固定大小的线程池，阻塞队列使用的是LinkedBlockingQueue

threadPool = Executors.newCachedThreadPool();//有缓冲的线程池，线程数 JVM 控制，阻塞队列使用的是SynchronousQueue

threadPool = Executors.newScheduledThreadPool(2);//定时任务功能的线程池

threadPool = new ThreadPoolExecutor();//默认线程池，可控制参数比较多

public static void main (String[] args) throws Exception {
    testThreadPoolExecutor();
}
public static void testThreadPoolExecutor() throws Exception {
    //基础参数
    int corePoolSize=2;//最小活跃线程数
    int maximumPoolSize=5;//最大活跃线程数
    int keepAliveTime=5;//指定线程池中线程空闲超过 5s 后将被回收
    TimeUnit unit = TimeUnit.SECONDS;//keepAliveTime 单位
    //阻塞队列
    BlockingQueue<Runnable> workQueue = null;
    workQueue = new ArrayBlockingQueue<>(5);//基于数组的先进先出队列，有界
    workQueue = new LinkedBlockingQueue<>();//基于链表的先进先出队列，无界
    workQueue = new SynchronousQueue<>();//无缓冲的等待队列，无界
    //拒绝策略
    RejectedExecutionHandler rejected = null;
    rejected = new ThreadPoolExecutor.AbortPolicy();//默认，队列满了丢任务抛出异常
    rejected = new ThreadPoolExecutor.DiscardPolicy();//队列满了丢任务不异常
    rejected = new ThreadPoolExecutor.DiscardOldestPolicy();//将最早进入队列的任务删，之后再尝试加入队列
    rejected = new ThreadPoolExecutor.CallerRunsPolicy();//如果添加到线程池失败，那么主线程会自己去执行该任务
    //使用的线程池
    ExecutorService threadPool = null;
    threadPool = Executors.newCachedThreadPool();//有缓冲的线程池，线程数 JVM 控制
    threadPool = Executors.newFixedThreadPool(3);//固定大小的线程池
    threadPool = Executors.newScheduledThreadPool(2);
    threadPool = Executors.newSingleThreadExecutor();//单线程的线程池，只有一个线程在工作
    threadPool = new ThreadPoolExecutor(
            corePoolSize,
            maximumPoolSize,
            keepAliveTime,
            unit,
            workQueue,
            rejected);//默认线程池，可控制参数比较多
    //执行无返回值线程
    TaskRunnable taskRunnable = new TaskRunnable();
    threadPool.execute(taskRunnable);
    List<Future<String>> futres = new ArrayList<>();
    for(int i=0;i<10;i++) {
        //执行有返回值线程
        TaskCallable taskCallable = new TaskCallable(i);
        Future<String> future = threadPool.submit(taskCallable);
        futres.add(future);
    }
    for(int i=0;i<futres.size();i++){
        String result = futres.get(i).get();
        System.out.println(i+" result = "+result);
    }
}
/**
    * 返回值的线程，使用 threadpool.execut() 执行
    */
public static class TaskRunnable implements Runnable{
    @Override
    public void run() {
        try {
            Thread.sleep(1000);
        } catch (InterruptedException e) {
            e.printStackTrace();
        }
        System.out.println(Thread.currentThread().getName() + " runnable result!");
    }
}
/**
    * 有返回值的线程，使用 threadpool.submit() 执行
    */
public static class TaskCallable implements Callable<String>{
    public TaskCallable(int index){
        this.i=index;
    }
    private int i;
    @Override
    public String call() throws Exception {
        int r = new Random().nextInt(5);
        try {
            Thread.sleep(r);
        } catch (InterruptedException e) {
            e.printStackTrace();
        }
        //System.out.println("callable result!");
        return Thread.currentThread().getName()+" callable index="+i +",sleep="+r;
    }
}