博客
关于我
强烈建议你试试无所不能的chatGPT,快点击我
ThreadPoolExecutor源码浅析
阅读量:5813 次
发布时间:2019-06-18

本文共 22494 字,大约阅读时间需要 74 分钟。

目录

初始化

ThreadPoolExecutor重载了多个构造方法,不过最终都是调用的同一个:

public ThreadPoolExecutor(int corePoolSize,                              int maximumPoolSize,                              long keepAliveTime,                              TimeUnit unit,                              BlockingQueue
workQueue, ThreadFactory threadFactory, RejectedExecutionHandler handler) { if (corePoolSize < 0 || maximumPoolSize <= 0 || maximumPoolSize < corePoolSize || keepAliveTime < 0) throw new IllegalArgumentException(); if (workQueue == null || threadFactory == null || handler == null) throw new NullPointerException(); this.acc = System.getSecurityManager() == null ? null : AccessController.getContext(); this.corePoolSize = corePoolSize; this.maximumPoolSize = maximumPoolSize; this.workQueue = workQueue; this.keepAliveTime = unit.toNanos(keepAliveTime); this.threadFactory = threadFactory; this.handler = handler; }

其中涉及了7个参数:

  • corePoolSize:线程池维护的线程数,及时线程空闲也不关闭,除非设置了allowCoreThreadTimeOut(默认未设置)
  • maximumPoolSize:最大线程数,当需要的线程数超过corePoolSize时就会新建线程,但线程总数不会超过maximumPoolSize
  • keepAliveTime:超出corePoolSize的线程,在用完后空闲时间超过keepAliveTime的时间后就会终止(terminating)
  • TimeUnit unit:keepAliveTime的时间单位
  • BlockingQueue<Runnable> workQueue:当任务无法立即被执行时,会被存储在队列中。不同类型的队列会导致线程池不同的特性,这里不深入讨论(有兴趣可以查看: 队列为 直接提交队列SynchronousQueue,无界队列LinkedBlockingQueue,有界队列ArrayBlockingQueue时不同的特性,)
  • ThreadFactory threadFactory:创建线程的工厂, 如常见的指定线程名字的工厂方法:new ThreadFactoryBuilder().setNameFormat("Thread-pool-%d").build();
  • RejectedExecutionHandler handler:拒绝策略,当线程数达到maximumPoolSize,且workQueue已经无法存储更多任务时,采用拒绝策略。

ThreadPoolExecutor为我们提供了4种拒绝策略:

  • AbortPolicy默认策略,抛出异常RejectedExecutionException,告诉调用方已经来不及处理了,调用方需要处理异常和线程线程池来不及执行的任务
  • DiscardPolicy,静默的忽略掉,无一致性要求的可以这么干
  • DiscardOldestPolicy,从队列里抛弃掉最老的任务,无一致性要求的可以这么干
  • CallerRunsPolicy,当任务添加到线程池中被拒绝时,会在线程池当前正在运行的Thread线程中处理被拒绝的任务。可以一定程度缓解当前线程不够的情况,但是如果当前任务执行所需时间不定,有卡住主线程的风险

再看看CallerRunsPolicy的实现:

public static class CallerRunsPolicy implements RejectedExecutionHandler {        /**         * Creates a {@code CallerRunsPolicy}.         */        public CallerRunsPolicy() { }        /**         * Executes task r in the caller's thread, unless the executor         * has been shut down, in which case the task is discarded.         *         * @param r the runnable task requested to be executed         * @param e the executor attempting to execute this task         */        public void rejectedExecution(Runnable r, ThreadPoolExecutor e) {            if (!e.isShutdown()) {                r.run();            }        }    }

可见是通过执行r.run()来占用主线程执行的。

所有的拒绝策略都是继承RejectedExecutionHandler,所以我们也可以自定义拒绝策略。

ctl变量

ctl变量是ThreadPoolExecutor的一个属性,ctl可以理解为control的简写,源码中定义如下:

private final AtomicInteger ctl = new AtomicInteger(ctlOf(RUNNING, 0));

源码中ctl变量的注释中解释了该变量的含义,该变量包含了两个含义,线程池的运行状态 (runState) 和线程池内有效线程的数量 (workerCount)。 ctl用高3位来表示线程池的运行状态, 用低29位来表示线程池内有效线程的数量。在源码中,rs通常表示线程池运行状态 , wc通常表示线程池中有效线程数量, 另外, ctl 也通常会简写作 c。

再看与ctl相关的几个变量和方法:

private static final int COUNT_BITS = Integer.SIZE - 3;    private static final int CAPACITY   = (1 << COUNT_BITS) - 1;    // runState is stored in the high-order bits    private static final int RUNNING    = -1 << COUNT_BITS;    private static final int SHUTDOWN   =  0 << COUNT_BITS;    private static final int STOP       =  1 << COUNT_BITS;    private static final int TIDYING    =  2 << COUNT_BITS;    private static final int TERMINATED =  3 << COUNT_BITS;    // Packing and unpacking ctl    private static int runStateOf(int c)     { return c & ~CAPACITY; }    private static int workerCountOf(int c)  { return c & CAPACITY; }    private static int ctlOf(int rs, int wc) { return rs | wc; }
  • COUNT_BITS,表示用于标记线程数量的位数,32-3=29位
  • CAPACITY, 表示线程池最大可以容纳的线程数量,2^30-1
  • RUNNING,表示运行状态,-1 << COUNT_BITS,前三位的值为111,后29位为0
  • SHUTDOWN,表示不接受新的任务,但是可以处理阻塞队列里的任务。0<< COUNT_BITS,前三位的值为000,后29位为0。调用shutdown()方法会置为该状态。
  • STOP,该状态不接受新的任务,不处理阻塞队列里的任务,中断正在处理的任务。1<< COUNT_BITS,前三位的值为001,后29位为0。调用shutdownNow()方法会置为该状态
  • TIDYING,表示过渡状态,2<< COUNT_BITS,前三位的值为010,后29位为0。此时表示所有的任务都执行完了,当前线程池已经没有有效的线程,并且将要调用terminated方法
  • TERMINATED,表示终止状态,3<< COUNT_BITS,前三位的值为011,后29位为0
  • runStateOf(int c) ,获取线程池状态,这里c为ctl变量,CAPACITY取反结果是前三位为1,后29位为0,与ctl与操作即可得到状态
  • workerCountOf(int c), 与runStateOf(int c) 相反取后29位,即线程数量
  • ctlOf(int rs, int wc),基于状态和线程数量构造一个ctl变量

对于状态可以简单理解为:RUNNING为-1,SHUTDOWN为0,STOP为1,TIDYING为2,TERMINATED为3。RUNNING变为SHUTDOWN或者STOP后,再变为TIDYING,再变为TERMINATED。

添加任务

ThreadPoolExecutor继承于AbstractExecutorService:

public class ThreadPoolExecutor extends AbstractExecutorService

AbstractExecutorService提供了最常用的三个添加任务到线程成的方法:

public Future
submit(Runnable task) { if (task == null) throw new NullPointerException(); RunnableFuture
ftask = newTaskFor(task, null); execute(ftask); return ftask; }public
Future
submit(Runnable task, T result) { if (task == null) throw new NullPointerException(); RunnableFuture
ftask = newTaskFor(task, result); execute(ftask); return ftask; }public
Future
submit(Callable
task) { if (task == null) throw new NullPointerException(); RunnableFuture
ftask = newTaskFor(task); execute(ftask); return ftask; }

可以看到最终它们都是调用了execute方法,ThreadPoolExecutor中execute的实现如下:

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.         *         * 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.         *         * 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.         */        int c = ctl.get();        if (workerCountOf(c) < corePoolSize) {            if (addWorker(command, true))                return;            c = ctl.get();        }        if (isRunning(c) && workQueue.offer(command)) {            int recheck = ctl.get();            if (! isRunning(recheck) && remove(command))                reject(command);            else if (workerCountOf(recheck) == 0)                addWorker(null, false);        }        else if (!addWorker(command, false))            reject(command);    }

源码中的这段注释详细的介绍了这段代码的作用,该方法考虑三种情况:

  1. 如果当前存活thread的数量小于corePoolSize,则尝试开启一个新的线程。如果创建成功则返回;如果创建失败,则继续后续步骤;
  2. 如果步骤1中创建失败或者thread数量>=corePoolSize,那会进入该步骤。该步骤判断线程池处于运行状态,则尝试将新任务加入队列。

    1. 如果线程池处于运行状态,且加入队列成功,则再次判断线程池是否处于运行状态(防止在执行workQueue.offer(command)的时候线程池状态改变)。如果线程池状态改变则remove刚刚入队的任务,并执行拒绝操作。如果在运行态,但是线程数为0,则添加一个worker。
    2. 如果线程池不处于运行状态加入队列失败则进入下一步骤
  3. 如果线程池不处于运行状态或者处于运行状态,但是thread数量>=corePoolSize且workQueue已满,则会进入该步骤。该步骤会尝试创建一个新的线程来执行任务。如果线程池线程总数达到maximumPoolSize 或者 创建线程时线程池状态变化不再处于运行状态,则会创建失败。

在上面的代码中主要是通过addWorker方法添加新任务的,下面我们就来分析下这个方法的实现

addWorker方法

源码如下:

private boolean addWorker(Runnable firstTask, boolean core) {        retry:        for (;;) {            int c = ctl.get();            int rs = runStateOf(c);            // Check if queue empty only if necessary.            //rs >= SHUTDOWN,状态不为RUNNING            //并且            //rs != SHUTDOWN || firstTask != null || workQueue.isEmpty()            //一下几种情况            //1. 状态不为RUNNING和SHUTDOWN,            //2. 或者 状态为SHUTDOWN且task不为null,            //3. 或者 状态为SHUTDOWN, task为null, workQueue 为空,            //则返回false,添加失败            if (rs >= SHUTDOWN &&                ! (rs == SHUTDOWN &&                   firstTask == null &&                   ! workQueue.isEmpty()))                return false;            //判断是否超过线程数量的限制,            for (;;) {                int wc = workerCountOf(c);                if (wc >= CAPACITY ||                    wc >= (core ? corePoolSize : maximumPoolSize))                    return false;                //未超过限制则尝试把线程数加1,成功跳出retry循环                if (compareAndIncrementWorkerCount(c))                    break retry;                //线程数加1失败则说明ctl有变化(状态或数量), 重新获取                c = ctl.get();  // Re-read ctl                //如果是状态变化则循环外层,反之循环内层                if (runStateOf(c) != rs)                     continue retry;                // else CAS failed due to workerCount change; retry inner loop            }        }        boolean workerStarted = false;        boolean workerAdded = false;        Worker w = null;        try {            w = new Worker(firstTask);            //从Worker构造方法可以看到             //this.firstTask = firstTask;            //this.thread = getThreadFactory().newThread(this);            //故此firstTask为null的时候, w.thread不为null            final Thread t = w.thread;            if (t != null) {                final ReentrantLock mainLock = this.mainLock;                mainLock.lock();                try {                    // Recheck while holding lock.                    // Back out on ThreadFactory failure or if                    // shut down before lock acquired.                    int rs = runStateOf(ctl.get());                    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();                }                if (workerAdded) {                    t.start();  //成功添加worker后,启动线程                    workerStarted = true;                }            }  //end of if (t != null)         } finally {            //worker启动失败则移除worker, 数量减一            if (! workerStarted)                addWorkerFailed(w);        }        return workerStarted;    }

execute方法中在三个地方用不用的参数调用了addWorker方法:

  1. addWorker(command, true)
  2. addWorker(null, false)
  3. addWorker(command, false)

addWorker有两个参数:Runnable firstTaskboolean core,前者表示要执行的任务,后者表示线程数量限制的类型(基于corePoolSize还是maximumPoolSize)。1和3 是类似的,唯一的不同就是线程数的限制不同,所以这里主要分析firstTask为null 和 不为null 的区别。

方法中retry: for (;;) {...}的内容主要是用于判断是否线程池已经关闭,以及线程数量是否超过限制。若未关闭,未超过限制则把线程数加1。firstTask为null的时候, w.thread不为null,所以firstTask是否在addWorker中还是没有区别,那只能更进一步看看worker里对firstTask是如何处理的。

worker实现

线程池中的任务都是通过worker来代理的。

private final class Worker        extends AbstractQueuedSynchronizer        implements Runnable{                /** Thread this worker is running in.  Null if factory fails. */        final Thread thread;        /** Initial task to run.  Possibly null. */        Runnable firstTask;        /** Per-thread task counter */        volatile long completedTasks;                //后续代码此处省略...........................}

Worker继承与AQS,实现Runable接口,本身是线程类,且具有AQS的特性。

看worker构造方法:

Worker(Runnable firstTask) {            setState(-1); // inhibit interrupts until runWorker            this.firstTask = firstTask;            //Worker实现了Runnable所以,            //所以this.thread.start(),就是用线程执行worker的run方法            this.thread = getThreadFactory().newThread(this);        }

setState(-1)为AQS的方法,把状态位设置成-1,这样任何线程都不能得到Worker的锁,除非调用了unlock方法。这个unlock方法会在runWorker方法中一开始就调用,这是为了确保Worker构造出来之后,没有任何线程能够得到它的锁,除非调用了runWorker之后,其他线程才能获得Worker的锁。

再看其run方法:

/** Delegates main run loop to outer runWorker  */        public void run() {            runWorker(this);        }

runWorker(this)不是worker的方法,是ThreadPoolExecutor的方法,也是执行任务的方法。

执行任务

又回到了ThreadPoolExecutor中,runWorker实现如下:

final void runWorker(Worker w) {        Thread wt = Thread.currentThread();        Runnable task = w.firstTask;        w.firstTask = null;        w.unlock(); // allow interrupts,创建worker时状态设置为-1了,此时设置为1        boolean completedAbruptly = true; //task是否意外终止,意外终止为true,反之false        try {            //优先运行初始化时的firstTask, 如果firstTask已经执行了则从队列取            while (task != null || (task = getTask()) != null) {                w.lock();  //获取到task后锁定,独占worker,保证线程安全                // If pool is stopping, ensure thread is interrupted;                // if not, ensure thread is not interrupted.  This                // requires a recheck in second case to deal with                // shutdownNow race while clearing interrupt                if ((runStateAtLeast(ctl.get(), STOP) ||                     (Thread.interrupted() &&                      runStateAtLeast(ctl.get(), STOP))) &&                    !wt.isInterrupted())                    wt.interrupt();                try {                    beforeExecute(wt, task); //空方法,用于子类扩展                    Throwable thrown = null;                    try {                        task.run();                    } catch (RuntimeException x) {                        thrown = x; throw x;                    } catch (Error x) {                        thrown = x; throw x;                    } catch (Throwable x) {                        thrown = x; throw new Error(x);                    } finally {                        afterExecute(task, thrown);//空方法,用于子类扩展                    }                } finally {                    task = null;                    w.completedTasks++;                    w.unlock();                }            }            completedAbruptly = false;        } finally {            //移除执行完成的worker            processWorkerExit(w, completedAbruptly);        }    }

到此我们终于能回答前面的问题了,addWorker(Runnable firstTask, boolean core) 中firstTask为null不不为null的区别:

  • 为null,addWorker(null, core) 表示创建一个worker,执行队列中的task
  • 不为null,addWorker(firstTask, core) 表示创建一个worker,先执行firstTask,再执行队列中的task

他们都新增了一个线程,一个是直接执行队列里的任务,一个先执行当前任务,再执行队列任务。

下面继续分析runWorker。

线程池在runWorker方法中,通过while (task != null || (task = getTask()) != null)不断从队列中取出任务执行,等待队列中任务执行完成后,调用processWorkerExit(w, completedAbruptly),移除当前worker。问题来了,这么看起来线程池中的线程只有在队列不为空的时候才得以复用,这不科学啊,那问题在哪儿?反复看代码,唯一忽略的掉的地方就是getTask()了,看到这个方法的时候,想当然的认为是简单的获取队列中的任务,那么我们来看一下它的具体实现:

private Runnable getTask() {        boolean timedOut = false; // Did the last poll() time out?        for (;;) {            int c = ctl.get();            int rs = runStateOf(c);            // Check if queue empty only if necessary. 线程池是否已经关闭            if (rs >= SHUTDOWN && (rs >= STOP || workQueue.isEmpty())) {                decrementWorkerCount();                return null;            }            int wc = workerCountOf(c);            // Are workers subject to culling?            //表示worker是否需要回收            //allowCoreThreadTimeOut=true时core线程超时也回收, 默认为false            //所以默认情况下timed表示 wc > corePoolSize            boolean timed = allowCoreThreadTimeOut || wc > corePoolSize;            if ((wc > maximumPoolSize || (timed && timedOut))                && (wc > 1 || workQueue.isEmpty())) {                if (compareAndDecrementWorkerCount(c))                    return null;                continue;            }            try {                Runnable r = timed ?                    //线程需要回收;尝试取队列中的任务,超过keepAliveTime还未取到返回null                    workQueue.poll(keepAliveTime, TimeUnit.NANOSECONDS) :                    //线程无需回收;取队列中的任务, 队列中没有任务则一直等到有任务                    workQueue.take();                if (r != null)                    return r;                timedOut = true;            } catch (InterruptedException retry) {                timedOut = false;            }        }    }

上面代码可以看出getTask()确实是取任务,不过也兼任了 线程池在运行态取不到数据时 park线程等待线程直到超时(parkNanos) 的工作,我们查看线程无需回收时park在取队列任务的线程堆栈如下:

"pool-1-thread-1@731" prio=5 tid=0xd nid=NA waiting  java.lang.Thread.State: WAITING      at sun.misc.Unsafe.park(Unsafe.java:-1)      at java.util.concurrent.locks.LockSupport.park(LockSupport.java:175)      at java.util.concurrent.locks.AbstractQueuedSynchronizer$ConditionObject.await(AbstractQueuedSynchronizer.java:2039)      at java.util.concurrent.ArrayBlockingQueue.take(ArrayBlockingQueue.java:403)      at java.util.concurrent.ThreadPoolExecutor.getTask(ThreadPoolExecutor.java:1074)      at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1134)      at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)      at java.lang.Thread.run(Thread.java:748)

线程处于waiting状态,从堆栈中可以看到at java.util.concurrent.ArrayBlockingQueue.take(ArrayBlockingQueue.java:403),正是被workQueue.take() park住了。如此一来worker执行完当前线程之后,如果取不到新的任务就会一直处在park状态,直到队列中有新的任务进入。以ArrayBlockingQueue为例看,看其takeenqueue实现:

/** Condition for waiting takes */ private final Condition notEmpty;public E take() throws InterruptedException {        final ReentrantLock lock = this.lock;        lock.lockInterruptibly();        try {            while (count == 0)                notEmpty.await();   //park 线程            return dequeue();        } finally {            lock.unlock();        }    }private void enqueue(E x) {        // assert lock.getHoldCount() == 1;        // assert items[putIndex] == null;        final Object[] items = this.items;        items[putIndex] = x;        if (++putIndex == items.length)            putIndex = 0;        count++;        notEmpty.signal();  //唤起线程    }

关闭连接池

ThreadPoolExecutor提供了两个关闭的方法:

  • shutdown(),关闭线程池,不再接受新的任务,但是会处理完当前线程和队列中的线程
  • shutdownNow() ,关闭线程池,不再接受新的任务,且试图停止所有正在执行的线程,并不再处理还在池队列中等待的任务。但是它试图终止线程的方法是通过调用Thread.interrupt()方法来实现的,但是interrupt的作用有限,运行中的线程不一定能成功退出(具体原因)。

下面看下实现:

public void shutdown() {        final ReentrantLock mainLock = this.mainLock;        mainLock.lock();        try {            checkShutdownAccess();            advanceRunState(SHUTDOWN);  //状态设置为SHUTDOWN            interruptIdleWorkers();  //中断空闲线程            onShutdown(); // hook for ScheduledThreadPoolExecutor,这里为空方法        } finally {            mainLock.unlock();        }        tryTerminate();    }    public List
shutdownNow() { List
tasks; final ReentrantLock mainLock = this.mainLock; mainLock.lock(); try { checkShutdownAccess(); advanceRunState(STOP); //状态设置为STOP interruptWorkers(); //中断全部线程 tasks = drainQueue(); //返回队列中未执行的任务 } finally { mainLock.unlock(); } tryTerminate(); return tasks; }

可以看到shutdown和shutdownNow的实现大致相同,不同的地方有两个,

  • 前者关闭时将状态设置为SHUTDOWN,后者为STOP
  • 前者interruptIdleWorkers(),只中断空闲线程;后者interruptWorkers(),中断全部 线程,返回队列中未执行的任务

设置状态的源码:

private void advanceRunState(int targetState) {        for (;;) {            int c = ctl.get();            if (runStateAtLeast(c, targetState) ||                ctl.compareAndSet(c, ctlOf(targetState, workerCountOf(c))))                break;        }    }

interruptIdleWorkers():

private void interruptIdleWorkers() {        interruptIdleWorkers(false);    }private void interruptIdleWorkers(boolean onlyOne) {        final ReentrantLock mainLock = this.mainLock;        mainLock.lock();        try {            for (Worker w : workers) {                Thread t = w.thread;                //如果线程未被中断,且获取work的锁成功(说明空闲),则中断线程                if (!t.isInterrupted() && w.tryLock()) {                    try {                        t.interrupt();                    } catch (SecurityException ignore) {                    } finally {                        w.unlock();                    }                }                if (onlyOne)                    break;            }        } finally {            mainLock.unlock();        }    }

interruptWorkers():

//ThreadPoolExecutorprivate void interruptWorkers() {        final ReentrantLock mainLock = this.mainLock;        mainLock.lock();        try {            //中断全部worker线程            for (Worker w : workers)                w.interruptIfStarted();        } finally {            mainLock.unlock();        }    }//workervoid interruptIfStarted() {            Thread t;            //若worker已经启动(未启动时为-1),且thread不为null,且未被中断            //也就是说线程还存活着,那就发送中断信号            if (getState() >= 0 && (t = thread) != null && !t.isInterrupted()) {                try {                    t.interrupt();                } catch (SecurityException ignore) {                }            }        }

tryTerminate()除了在关闭连接池时调用,还在其它地方调用了,这里只分析在关闭连接池时它都做了什么:

final void tryTerminate() {        for (;;) {            int c = ctl.get();            //关闭连接池调用该方法第一次调用时:            //状态为SHUTDOWN或STOP,都小于TIDYING,故前两条件都不满足            //第三个条件,队列不为空的时候直接返回了,            //如果为shutdown()则可能队列不为空,可能满足条件直接返回,也可能不满足            //如果为shutdownNow()则队列被清空,不满足            if (isRunning(c) ||                runStateAtLeast(c, TIDYING) ||                (runStateOf(c) == SHUTDOWN && ! workQueue.isEmpty()))                return;            //如果worker数量不为0则执行interruptIdleWorkers(true)            //然后直接返回,完成该方法            if (workerCountOf(c) != 0) { // Eligible to terminate                interruptIdleWorkers(ONLY_ONE);                return;            }            final ReentrantLock mainLock = this.mainLock;            mainLock.lock();            try {                //尝试设置状态为TIDYING,worker数量为0,                //期间ctl若未变动,则成功                if (ctl.compareAndSet(c, ctlOf(TIDYING, 0))) {                    try {                        terminated(); //空方法用于子类扩展                    } finally {                        //设置状态为TERMINATED                        ctl.set(ctlOf(TERMINATED, 0));                        //唤醒调用了awaitTermination(long timeout, TimeUnit unit)的线程                        //awaitTermination中调用了                        termination.signalAll();                    }                    return;                }            } finally {                mainLock.unlock();            }            // else retry on failed CAS        }    }

tryTerminate()在关闭连接池时的做的判断可以简单理解为

    • 如果队列不为空直接返回
    • 存活worker数量不为0则直接返回
    • 设置状态为TIDYING,TERMINATED

所以无论是shutdown还是shutdownNow都不会阻塞线程,且不保证worker已经全部关闭。

参考

转载于:https://www.cnblogs.com/chrischennx/p/9600156.html

你可能感兴趣的文章
oracle import & export 操作相关脚本
查看>>
LVS集群的体系结构,构建强壮的体系结构里负载均衡层、真实服务器层、后端共享存储层都是相辅相成...
查看>>
VTK读取raw格式的数据
查看>>
java泛型
查看>>
J2EE搭建之十一 如何设置start-on-load
查看>>
像这样的高额授权费在业界并不常见
查看>>
我的友情链接
查看>>
推荐几本书
查看>>
批量实现多台服务器之间ssh无密码登录的相互信任关系
查看>>
思科ACL详解
查看>>
mysql自动备份脚本及异地定时FTP
查看>>
使用Django开发一个图书管理系统------02-需求分析设计
查看>>
PMP知识要点(第六章)
查看>>
设计模式-Composite Pattern
查看>>
彻底明白IP地址——计算相关地址
查看>>
Vmware Workstation 安装Oracle Solaris 10
查看>>
《统一沟通-微软-实战》-7-配置-1-电话拨入式会议
查看>>
MySQL 5.7.17 Group Replication搭建
查看>>
手把手教你跑Larave框架实战笔记系列之二
查看>>
Keepalived双主模型实现nginx负载均衡
查看>>