Java线程池理解及用法

前言

多线程的异步执行方式，虽然能够最大限度发挥多核计算机的计算能力，但是如果不加控制，反而会对系统造成负担。线程本身也要占用内存空间，大量的线程会占用内存资源并且可能会导致Out of Memory。即便没有这样的情况，大量的线程回收也会给GC带来很大的压力。

为了避免重复的创建线程，线程池的出现可以让线程进行复用。通俗点讲，当有工作来，就会向线程池拿一个线程，当工作完成后，并不是直接关闭线程，而是将这个线程归还给线程池供其他任务使用。

接下来从总体到细致的方式，来共同探讨线程池。

总体的架构

来看Executor的框架图：

aaarticlea/png;base64,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" alt="" />

接口：Executor,CompletionService,ExecutorService，ScheduledExecutorService

抽象类：AbstractExecutorService

实现类：ExecutorCompletionService，ThreadPoolExecutor，ScheduledThreadPoolExecutor

从图中就可以看到主要的方法，本文主要讨论的是ThreadPoolExecutor

研读ThreadPoolExecutor

看一下该类的构造器：

    public ThreadPoolExecutor(int paramInt1, int paramInt2, long paramLong, TimeUnit paramTimeUnit,
            BlockingQueue<Runnable> paramBlockingQueue, ThreadFactory paramThreadFactory,
            RejectedExecutionHandler paramRejectedExecutionHandler) {
        this.ctl = new AtomicInteger(ctlOf(-536870912, 0));
        this.mainLock = new ReentrantLock();
        this.workers = new HashSet();
        this.termination = this.mainLock.newCondition();
        if ((paramInt1 < 0) || (paramInt2 <= 0) || (paramInt2 < paramInt1) || (paramLong < 0L))
            throw new IllegalArgumentException();
        if ((paramBlockingQueue == null) || (paramThreadFactory == null) || (paramRejectedExecutionHandler == null))
            throw new NullPointerException();
        this.corePoolSize = paramInt1;
        this.maximumPoolSize = paramInt2;
        this.workQueue = paramBlockingQueue;
        this.keepAliveTime = paramTimeUnit.toNanos(paramLong);
        this.threadFactory = paramThreadFactory;
        this.handler = paramRejectedExecutionHandler;
    }

corePoolSize :线程池的核心池大小，在创建线程池之后，线程池默认没有任何线程。

当有任务过来的时候才会去创建创建线程执行任务。换个说法，线程池创建之后，线程池中的线程数为0，当任务过来就会创建一个线程去执行，直到线程数达到corePoolSize 之后，就会被到达的任务放在队列中。（注意是到达的任务）。换句更精炼的话：corePoolSize 表示允许线程池中允许同时运行的最大线程数。

如果执行了线程池的prestartAllCoreThreads()方法，线程池会提前创建并启动所有核心线程。

maximumPoolSize :线程池允许的最大线程数，他表示最大能创建多少个线程。maximumPoolSize肯定是大于等于corePoolSize。

keepAliveTime :表示线程没有任务时最多保持多久然后停止。默认情况下，只有线程池中线程数大于corePoolSize 时，keepAliveTime 才会起作用。换句话说，当线程池中的线程数大于corePoolSize，并且一个线程空闲时间达到了keepAliveTime，那么就是shutdown。

Unit:keepAliveTime 的单位。

workQueue ：一个阻塞队列，用来存储等待执行的任务，当线程池中的线程数超过它的corePoolSize的时候，线程会进入阻塞队列进行阻塞等待。通过workQueue，线程池实现了阻塞功能

threadFactory ：线程工厂，用来创建线程。

handler :表示当拒绝处理任务时的策略。

任务缓存队列

在前面我们多次提到了任务缓存队列，即workQueue，它用来存放等待执行的任务。

workQueue的类型为BlockingQueue<Runnable>，通常可以取下面三种类型：

1）有界任务队列ArrayBlockingQueue：基于数组的先进先出队列，此队列创建时必须指定大小；

2）无界任务队列LinkedBlockingQueue：基于链表的先进先出队列，如果创建时没有指定此队列大小，则默认为Integer.MAX_VALUE；

3）直接提交队列synchronousQueue：这个队列比较特殊，它不会保存提交的任务，而是将直接新建一个线程来执行新来的任务。

拒绝策略

AbortPolicy:丢弃任务并抛出RejectedExecutionException

CallerRunsPolicy：只要线程池未关闭，该策略直接在调用者线程中，运行当前被丢弃的任务。显然这样做不会真的丢弃任务，但是，任务提交线程的性能极有可能会急剧下降。

DiscardOldestPolicy：丢弃队列中最老的一个请求，也就是即将被执行的一个任务，并尝试再次提交当前任务。

DiscardPolicy：丢弃任务，不做任何处理。

线程池的任务处理策略：

如果当前线程池中的线程数目小于corePoolSize，则每来一个任务，就会创建一个线程去执行这个任务；

如果当前线程池中的线程数目>=corePoolSize，则每来一个任务，会尝试将其添加到任务缓存队列当中，若添加成功，则该任务会等待空闲线程将其取出去执行；若添加失败（一般来说是任务缓存队列已满），则会尝试创建新的线程去执行这个任务；如果当前线程池中的线程数目达到maximumPoolSize，则会采取任务拒绝策略进行处理；

如果线程池中的线程数量大于 corePoolSize时，如果某线程空闲时间超过keepAliveTime，线程将被终止，直至线程池中的线程数目不大于corePoolSize；如果允许为核心池中的线程设置存活时间，那么核心池中的线程空闲时间超过keepAliveTime，线程也会被终止。

线程池的关闭

ThreadPoolExecutor提供了两个方法，用于线程池的关闭，分别是shutdown()和shutdownNow()，其中：

shutdown()：不会立即终止线程池，而是要等所有任务缓存队列中的任务都执行完后才终止，但再也不会接受新的任务

shutdownNow()：立即终止线程池，并尝试打断正在执行的任务，并且清空任务缓存队列，返回尚未执行的任务

源码分析

首先来看最核心的execute方法，这个方法在AbstractExecutorService中并没有实现，从Executor接口，直到ThreadPoolExecutor才实现了改方法，

ExecutorService中的submit(),invokeAll(),invokeAny()都是调用的execute方法，所以execute是核心中的核心,源码分析将围绕它逐步展开。

    public void execute(Runnable command) {
        if (command == null)
            throw new NullPointerException();
        /*
         * Proceed in 3 steps:
         *
         * 1. If fewer than corePoolSize threads are running, try to
         * start a new thread with the given command as its first
         * task.  The call to addWorker atomically checks runState and
         * workerCount, and so prevents false alarms that would add
         * threads when it shouldn't, by returning false.
         * 如果正在运行的线程数小于corePoolSize，那么将调用addWorker 方法来创建一个新的线程，并将该任务作为新线程的第一个任务来执行。
　　　　　　 当然，在创建线程之前会做原子性质的检查，如果条件不允许，则不创建线程来执行任务，并返回false.　　
         * 2. If a task can be successfully queued, then we still need
         * to double-check whether we should have added a thread
         * (because existing ones died since last checking) or that
         * the pool shut down since entry into this method. So we
         * recheck state and if necessary roll back the enqueuing if
         * stopped, or start a new thread if there are none.
         * 如果一个任务成功进入阻塞队列，那么我们需要进行一个双重检查来确保是我们已经添加一个线程（因为存在着一些线程在上次检查后他已经死亡）或者
　　　　　　 当我们进入该方法时，该线程池已经关闭。所以，我们将重新检查状态，线程池关闭的情况下则回滚入队列，线程池没有线程的情况则创建一个新的线程。
         * 3. If we cannot queue task, then we try to add a new
         * thread.  If it fails, we know we are shut down or saturated
         * and so reject the task.
　　　　　　 如果任务无法入队列（队列满了），那么我们将尝试新开启一个线程（从corepoolsize到扩充到maximum），如果失败了，那么可以确定原因，要么是
　　　　　　 线程池关闭了或者饱和了（达到maximum），所以我们执行拒绝策略。
         */
　　　　
　　　　// 1.当前线程数量小于corePoolSize，则创建并启动线程。
        int c = ctl.get();
        if (workerCountOf(c) < corePoolSize) {
            if (addWorker(command, true))
　　　　　　　　// 成功，则返回

return;
            c = ctl.get();
        }
　　　　// 2.步骤1失败，则尝试进入阻塞队列，
        if (isRunning(c) && workQueue.offer(command)) {
　　　　　　　// 入队列成功，检查线程池状态，如果状态部署RUNNING而且remove成功，则拒绝任务
            int recheck = ctl.get();
            if (! isRunning(recheck) && remove(command))
                reject(command);
　　　　　　　// 如果当前worker数量为0，通过addWorker(null, false)创建一个线程，其任务为null
            else if (workerCountOf(recheck) == 0)
                addWorker(null, false);
        }
　　　　// 3. 步骤1和2失败，则尝试将线程池的数量有corePoolSize扩充至maxPoolSize，如果失败，则拒绝任务
        else if (!addWorker(command, false))
            reject(command);
    }

相信看了代码也是一脸懵,接下来用一个流程图来讲一讲，他究竟干了什么事：

结合上面的流程图来逐行解析，首先前面进行空指针检查，

wonrkerCountOf()方法能够取得当前线程池中的线程的总数，取得当前线程数与核心池大小比较，

• 如果小于，将通过addWorker()方法调度执行。
• 如果大于核心池大小，那么就提交到等待队列。
• 如果进入等待队列失败，则会将任务直接提交给线程池。
• 如果线程数达到最大线程数，那么就提交失败，执行拒绝策略。

excute()方法中添加任务的方式是使用addWorker（）方法，看一下源码，一起学习一下。

private boolean addWorker(Runnable firstTask, boolean core) {
        retry:
　　　　 // 外层循环，用于判断线程池状态
        for (;;) {
            int c = ctl.get();
            int rs = runStateOf(c);

            // Check if queue empty only if necessary.
            if (rs >= SHUTDOWN &&
                ! (rs == SHUTDOWN &&
                   firstTask == null &&
                   ! workQueue.isEmpty()))
                return false;
　　　　　　 // 内层的循环，任务是将worker数量加1
            for (;;) {
                int wc = workerCountOf(c);
                if (wc >= CAPACITY ||
                    wc >= (core ? corePoolSize : maximumPoolSize))
                    return false;
                if (compareAndIncrementWorkerCount(c))
                    break retry;
                c = ctl.get();  // Re-read ctl
                if (runStateOf(c) != rs)
                    continue retry;
                // else CAS failed due to workerCount change; retry inner loop
            }
        }
　　　　// worker加1后，接下来将woker添加到HashSet<Worker>中，并启动worker
        boolean workerStarted = false;
        boolean workerAdded = false;
        Worker w = null;
        try {
            final ReentrantLock mainLock = this.mainLock;
            w = new Worker(firstTask);
            final Thread t = w.thread;
            if (t != null) {
                mainLock.lock();
                try {
                    // Recheck while holding lock.
                    // Back out on ThreadFactory failure or if
                    // shut down before lock acquired.
                    int c = ctl.get();
                    int rs = runStateOf(c);

                    if (rs < SHUTDOWN ||
                        (rs == SHUTDOWN && firstTask == null)) {
                        if (t.isAlive()) // precheck that t is startable
                            throw new IllegalThreadStateException();
                        workers.add(w);
                        int s = workers.size();
                        if (s > largestPoolSize)
                            largestPoolSize = s;
                        workerAdded = true;
                    }
                } finally {
                    mainLock.unlock();
                }
　　　　　　　　　// 如果往HashSet<Worker>添加成功，则启动该线程
                if (workerAdded) {
                    t.start();
                    workerStarted = true;
                }
            }
        } finally {
            if (! workerStarted)
                addWorkerFailed(w);
        }
        return workerStarted;
    }

addWorker(Runnable firstTask, boolean core)的主要任务是创建并启动线程。

他会根据当前线程的状态和给定的值（core or maximum）来判断是否可以创建一个线程。

addWorker共有四种传参方式。execute使用了其中三种，分别为:

1.addWorker(paramRunnable, true)

线程数小于corePoolSize时，放一个需要处理的task进Workers Set。如果Workers Set长度超过corePoolSize，就返回false.

2.addWorker(null, false)

放入一个空的task进workers Set，长度限制是maximumPoolSize。这样一个task为空的worker在线程执行的时候会去任务队列里拿任务，这样就相当于创建了一个新的线程，只是没有马上分配任务。

3.addWorker(paramRunnable, false)

当队列被放满时，就尝试将这个新来的task直接放入Workers Set，而此时Workers Set的长度限制是maximumPoolSize。如果线程池也满了的话就返回false.

还有一种情况是execute()方法没有使用的

addWorker(null, true)

这个方法就是放一个null的task进Workers Set，而且是在小于corePoolSize时，如果此时Set中的数量已经达到corePoolSize那就返回false，什么也不干。实际使用中是在prestartAllCoreThreads()方法，这个方法用来为线程池预先启动corePoolSize个worker等待从workQueue中获取任务执行。

执行流程：

1、判断线程池当前是否为可以添加worker线程的状态，可以则继续下一步，不可以return false：
    A、线程池状态>shutdown，可能为stop、tidying、terminated，不能添加worker线程
    B、线程池状态==shutdown，firstTask不为空，不能添加worker线程，因为shutdown状态的线程池不接收新任务
    C、线程池状态==shutdown，firstTask==null，workQueue为空，不能添加worker线程，因为firstTask为空是为了添加一个没有任务的线程再从workQueue获取task，而workQueue为  　　　　空，说明添加无任务线程已经没有意义
2、线程池当前线程数量是否超过上限（corePoolSize 或 maximumPoolSize），超过了return false，没超过则对workerCount+1，继续下一步
3、在线程池的ReentrantLock保证下，向Workers Set中添加新创建的worker实例，添加完成后解锁，并启动worker线程，如果这一切都成功了，return true，如果添加worker入Set失败或启动失败，调用addWorkerFailed()逻辑

常见的四种线程池

newFixedThreadPool

public static ExecutorService newFixedThreadPool(int var0) {
        return new ThreadPoolExecutor(var0, var0, 0L, TimeUnit.MILLISECONDS, new LinkedBlockingQueue());
    }

public static ExecutorService newFixedThreadPool(int var0, ThreadFactory var1) {
    return new ThreadPoolExecutor(var0, var0, 0L, TimeUnit.MILLISECONDS, new LinkedBlockingQueue(), var1);
}

固定大小的线程池，可以指定线程池的大小，该线程池corePoolSize和maximumPoolSize相等，阻塞队列使用的是LinkedBlockingQueue，大小为整数最大值。

该线程池中的线程数量始终不变，当有新任务提交时，线程池中有空闲线程则会立即执行，如果没有，则会暂存到阻塞队列。对于固定大小的线程池，不存在线程数量的变化。同时使用无界的LinkedBlockingQueue来存放执行的任务。当任务提交十分频繁的时候，LinkedBlockingQueue

迅速增大，存在着耗尽系统资源的问题。而且在线程池空闲时，即线程池中没有可运行任务时，它也不会释放工作线程，还会占用一定的系统资源，需要shutdown。

newSingleThreadExecutor

public static ExecutorService newSingleThreadExecutor() {
        return new Executors.FinalizableDelegatedExecutorService(new ThreadPoolExecutor(1, 1, 0L, TimeUnit.MILLISECONDS, new LinkedBlockingQueue()));
    }

    public static ExecutorService newSingleThreadExecutor(ThreadFactory var0) {
        return new Executors.FinalizableDelegatedExecutorService(new ThreadPoolExecutor(1, 1, 0L, TimeUnit.MILLISECONDS, new LinkedBlockingQueue(), var0));
    }

单个线程线程池，只有一个线程的线程池，阻塞队列使用的是LinkedBlockingQueue,若有多余的任务提交到线程池中，则会被暂存到阻塞队列，待空闲时再去执行。按照先入先出的顺序执行任务。

newCachedThreadPool

public static ExecutorService newCachedThreadPool() {
        return new ThreadPoolExecutor(0, 2147483647, 60L, TimeUnit.SECONDS, new SynchronousQueue());
    }

    public static ExecutorService newCachedThreadPool(ThreadFactory var0) {
        return new ThreadPoolExecutor(0, 2147483647, 60L, TimeUnit.SECONDS, new SynchronousQueue(), var0);
    }

缓存线程池，缓存的线程默认存活60秒。线程的核心池corePoolSize大小为0，核心池最大为Integer.MAX_VALUE,阻塞队列使用的是SynchronousQueue。是一个直接提交的阻塞队列，    他总会迫使线程池增加新的线程去执行新的任务。在没有任务执行时，当线程的空闲时间超过keepAliveTime（60秒），则工作线程将会终止被回收，当提交新任务时，如果没有空闲线程，则创建新线程执行任务，会导致一定的系统开销。如果同时又大量任务被提交，而且任务执行的时间不是特别快，那么线程池便会新增出等量的线程池处理任务，这很可能会很快耗尽系统的资源。

newScheduledThreadPool

public static ScheduledExecutorService newScheduledThreadPool(int var0) {
        return new ScheduledThreadPoolExecutor(var0);
    }

    public static ScheduledExecutorService newScheduledThreadPool(int var0, ThreadFactory var1) {
        return new ScheduledThreadPoolExecutor(var0, var1);
    }

定时线程池，该线程池可用于周期性地去执行任务，通常用于周期性的同步数据。

scheduleAtFixedRate:是以固定的频率去执行任务，周期是指每次执行任务成功执行之间的间隔。

schedultWithFixedDelay:是以固定的延时去执行任务，延时是指上一次执行成功之后和下一次开始执行的之前的时间。

使用实例

newFixedThreadPool实例：

 

newCachedThreadPool实例：

 

这里没用调用shutDown方法，这里可以发现过60秒之后，会自动释放资源。

newSingleThreadExecutor

 

这里需要注意一点，newSingleThreadExecutor和newFixedThreadPool一样，在线程池中没有任务时可执行，也不会释放系统资源的，所以需要shudown。

newScheduledThreadPool

 

最后杂谈

如何选择线程池数量

线程池的大小决定着系统的性能，过大或者过小的线程池数量都无法发挥最优的系统性能。

当然线程池的大小也不需要做的太过于精确，只需要避免过大和过小的情况。一般来说，确定线程池的大小需要考虑CPU的数量，内存大小，任务是计算密集型还是IO密集型等因素

NCPU = CPU的数量

UCPU = 期望对CPU的使用率 0 ≤ UCPU ≤ 1

W/C = 等待时间与计算时间的比率

如果希望处理器达到理想的使用率，那么线程池的最优大小为：

线程池大小=NCPU *UCPU(1+W/C)

在Java中使用

int ncpus = Runtime.getRuntime().availableProcessors();

获取CPU的数量。

线程池工厂

Executors的线程池如果不指定线程工厂会使用Executors中的DefaultThreadFactory,默认线程池工厂创建的线程都是非守护线程。

使用自定义的线程工厂可以做很多事情，比如可以跟踪线程池在何时创建了多少线程，也可以自定义线程名称和优先级。如果将

新建的线程都设置成守护线程，当主线程退出后，将会强制销毁线程池。

下面这个例子，记录了线程的创建，并将所有的线程设置成守护线程。

 

扩展线程池

ThreadPoolExecutor是可以拓展的，它提供了几个可以在子类中改写的方法：beforeExecute,afterExecute和terimated。

在执行任务的线程中将调用beforeExecute和afterExecute,这些方法中还可以添加日志，计时，监视或统计收集的功能，

还可以用来输出有用的调试信息，帮助系统诊断故障。下面是一个扩展线程池的例子：

 

线程池的正确使用

以下阿里编码规范里面说的一段话：

线程池不允许使用Executors去创建，而是通过ThreadPoolExecutor的方式，这样的处理方式让写的同学更加明确线程池的运行规则，规避资源耗尽的风险。 说明：Executors各个方法的弊端：
1）newFixedThreadPool和newSingleThreadExecutor:
  主要问题是堆积的请求处理队列可能会耗费非常大的内存，甚至OOM。
2）newCachedThreadPool和newScheduledThreadPool:
  主要问题是线程数最大数是Integer.MAX_VALUE，可能会创建数量非常多的线程，甚至OOM。

手动创建线程池有几个注意点

1.任务独立。如何任务依赖于其他任务，那么可能产生死锁。例如某个任务等待另一个任务的返回值或执行结果，那么除非线程池足够大，否则将发生线程饥饿死锁。

2.合理配置阻塞时间过长的任务。如果任务阻塞时间过长，那么即使不出现死锁，线程池的性能也会变得很糟糕。在Java并发包里可阻塞方法都同时定义了限时方式和不限时方式。例如

Thread.join,BlockingQueue.put,CountDownLatch.await等，如果任务超时，则标识任务失败，然后中止任务或者将任务放回队列以便随后执行，这样，无论任务的最终结果是否成功，这种办法都能够保证任务总能继续执行下去。

3.设置合理的线程池大小。只需要避免过大或者过小的情况即可，上文的公式线程池大小=NCPU *UCPU(1+W/C)。

4.选择合适的阻塞队列。newFixedThreadPool和newSingleThreadExecutor都使用了无界的阻塞队列，无界阻塞队列会有消耗很大的内存，如果使用了有界阻塞队列，它会规避内存占用过大的问题，但是当任务填满有界阻塞队列，新的任务该怎么办？在使用有界队列是，需要选择合适的拒绝策略，队列的大小和线程池的大小必须一起调节。对于非常大的或者无界的线程池，可以使用SynchronousQueue来避免任务排队，以直接将任务从生产者提交到工作者线程。

下面是Thrift框架处理socket任务所使用的一个线程池，可以看一下FaceBook的工程师是如何自定义线程池的。

    private static ExecutorService createDefaultExecutorService(Args args) {
        SynchronousQueue executorQueue = new SynchronousQueue();

        return new ThreadPoolExecutor(args.minWorkerThreads, args.maxWorkerThreads, 60L, TimeUnit.SECONDS,
                executorQueue);
    }

TEST

新建线程池：
   protected ExecutorService testThreadPool = Executors.newFixedThreadPool(5);

   Map<String, String> scoreTestMap = new HashMap<String, String>();

   TestCallable testCallable = new TestCallable(testService, words, clusRedisTemplate);
   FutureTask<Map<String,String>> futureTestTask = new FutureTask<Map<String,String>>(testCallable); 
   testThreadPool .submit(futureTestTask);
   scoreTestMap = futureTestTask.get(200, TimeUnit.MILLISECONDS);

总结：

查看以备份

参考资料

《实战Java》高并发程序设计

《Java Concurrency in Practice》

Java线程池ThreadPoolExecutor使用和分析(二)

https://www.jb51.net/article/121296.htm

https://blog.csdn.net/u011277123/article/details/78257336

多线程使用场景：

1. 常见的浏览器、Web服务(现在写的web是中间件帮你完成了线程的控制)，web处理请求，各种专用服务器(如游戏服务器)
2. servlet多线程
3. FTP下载，多线程操作文件
4. 数据库用到的多线程
5. 分布式计算
6. tomcat，tomcat内部采用多线程，上百个客户端访问同一个WEB应用，tomcat接入后就是把后续的处理扔给一个新的线程来处理，这个新的线程最后调用我们的servlet程序，比如doGet或者dpPost方法
7. 后台任务：如定时向大量(100W以上)的用户发送邮件；定期更新配置文件、任务调度(如quartz)，一些监控用于定期信息采集
8. 自动作业处理：比如定期备份日志、定期备份数据库
9. 异步处理：如发微博、记录日志
10. 页面异步处理：比如大批量数据的核对工作(有10万个手机号码，核对哪些是已有用户)
11. 数据库的数据分析(待分析的数据太多)，数据迁移
12. 多步骤的任务处理，可根据步骤特征选用不同个数和特征的线程来协作处理，多任务的分割，由一个主线程分割给多个线程完成
13. desktop应用开发，一个费时的计算开个线程，前台加个进度条显示
14. swing编程