Java源码里面都有大量的注释，认真读懂这些注释，就可以把握其七分工作机制了。关于ThreadPoolExecutor的解析，我们就从其类注释开始。

ThreadPoolExecutor.png

现将注释大致翻译如下：

ExecutorService（ThreadPoolExecutor的顶层接口）使用线程池中的线程执行每个提交的任务，通常我们使用Executors的工厂方法来创建ExecutorService。

线程池解决了两个不同的问题：

1. 提升性能：它们通常在执行大量异步任务时，由于减少了每个任务的调用开销，并且它们提供了一种限制和管理资源（包括线程）的方法，使得性能提升明显；
2. 统计信息：每个ThreadPoolExecutor保持一些基本的统计信息，例如完成的任务数量。

为了在广泛的上下文中有用，此类提供了许多可调参数和可扩展性钩子。 但是，在常见场景中，我们预配置了几种线程池，我们敦促程序员使用更方便的Executors的工厂方法直接使用。

• Executors.newCachedThreadPool（无界线程池，自动线程回收）
• Executors.newFixedThreadPool（固定大小的线程池）；
• Executors.newSingleThreadExecutor（单一后台线程）；

注：这里没有提到ScheduledExecutorService ，后续解析。

在自定义线程池时，请参考以下指南：

一、Core and maximum pool sizes 核心和最大线程池数量

线程池执行器将会根据corePoolSize和maximumPoolSize自动地调整线程池大小。

当在execute(Runnable)方法中提交新任务并且少于corePoolSize线程正在运行时，即使其他工作线程处于空闲状态，也会创建一个新线程来处理该请求。 如果有多于corePoolSize但小于maximumPoolSize线程正在运行，则仅当队列已满时才会创建新线程。 通过设置corePoolSize和maximumPoolSize相同，您可以创建一个固定大小的线程池。 通过将maximumPoolSize设置为基本上无界的值，例如Integer.MAX_VALUE，您可以允许池容纳任意数量的并发任务。 通常，核心和最大池大小仅在构建时设置，但也可以使用setCorePoolSize和setMaximumPoolSize进行动态更改。

这段话详细了描述了线程池对任务的处理流程，这里用个图总结一下

线程任务处理流程.png

二、prestartCoreThread 核心线程预启动
在默认情况下，只有当新任务到达时，才开始创建和启动核心线程，但是我们可以使用 prestartCoreThread() 和 prestartAllCoreThreads() 方法动态调整。
如果使用非空队列构建池，则可能需要预先启动线程。

三、ThreadFactory 线程工厂

新线程使用ThreadFactory创建。 如果未另行指定，则使用Executors.defaultThreadFactory默认工厂，使其全部位于同一个ThreadGroup中，并且具有相同的NORM_PRIORITY优先级和非守护进程状态。

通过提供不同的ThreadFactory，您可以更改线程的名称，线程组，优先级，守护进程状态等。如果ThreadCactory在通过从newThread返回null询问时未能创建线程，则执行程序将继续，但可能无法执行任何任务。

线程应该有modifyThread权限。 如果工作线程或使用该池的其他线程不具备此权限，则服务可能会降级：配置更改可能无法及时生效，并且关闭池可能会保持可终止但尚未完成的状态。

四、Keep-alive times 线程存活时间

如果线程池当前拥有超过corePoolSize的线程，那么多余的线程在空闲时间超过keepAliveTime时会被终止 ( 请参阅getKeepAliveTime(TimeUnit) )。这提供了一种在不积极使用线程池时减少资源消耗的方法。

如果池在以后变得更加活跃，则应构建新线程。 也可以使用方法setKeepAliveTime(long，TimeUnit)进行动态调整。

防止空闲线程在关闭之前终止，可以使用如下方法：

setKeepAliveTime(Long.MAX_VALUE，TimeUnit.NANOSECONDS);

默认情况下，keep-alive策略仅适用于存在超过corePoolSize线程的情况。 但是，只要keepAliveTime值不为零，方法allowCoreThreadTimeOut(boolean)也可用于将此超时策略应用于核心线程。

五、Queuing 队列

BlockingQueu用于存放提交的任务，队列的实际容量与线程池大小相关联。

• 如果当前线程池任务线程数量小于核心线程池数量，执行器总是优先创建一个任务线程，而不是从线程队列中取一个空闲线程。

• 如果当前线程池任务线程数量大于核心线程池数量，执行器总是优先从线程队列中取一个空闲线程，而不是创建一个任务线程。

• 如果当前线程池任务线程数量大于核心线程池数量，且队列中无空闲任务线程，将会创建一个任务线程，直到超出maximumPoolSize，如果超时maximumPoolSize，则任务将会被拒绝。

这个过程参考[线程任务处理流程图.png]

主要有三种队列策略：

1. Direct handoffs 直接握手队列
   Direct handoffs 的一个很好的默认选择是 SynchronousQueue，它将任务交给线程而不需要保留。这里，如果没有线程立即可用来运行它，那么排队任务的尝试将失败，因此将构建新的线程。
   此策略在处理可能具有内部依赖关系的请求集时避免锁定。Direct handoffs 通常需要无限制的maximumPoolSizes来避免拒绝新提交的任务。 但得注意，当任务持续以平均提交速度大余平均处理速度时，会导致线程数量会无限增长问题。

2. Unbounded queues 无界队列
   当所有corePoolSize线程繁忙时，使用无界队列（例如，没有预定义容量的LinkedBlockingQueue）将导致新任务在队列中等待，从而导致maximumPoolSize的值没有任何作用。当每个任务互不影响，完全独立于其他任务时，这可能是合适的; 例如，在网页服务器中， 这种队列方式可以用于平滑瞬时大量请求。但得注意，当任务持续以平均提交速度大余平均处理速度时，会导致队列无限增长问题。

3. Bounded queues 有界队列
   一个有界的队列（例如，一个ArrayBlockingQueue）和有限的maximumPoolSizes配置有助于防止资源耗尽，但是难以控制。队列大小和maximumPoolSizes需要 相互权衡：

• 使用大队列和较小的maximumPoolSizes可以最大限度地减少CPU使用率，操作系统资源和上下文切换开销，但会导致人为的低吞吐量。如果任务经常被阻塞（比如I/O限制），那么系统可以调度比我们允许的更多的线程。
• 使用小队列通常需要较大的maximumPoolSizes，这会使CPU更繁忙，但可能会遇到不可接受的调度开销，这也会降低吞吐量。
  这里主要为了说明有界队列大小和maximumPoolSizes的大小控制，若何降低资源消耗的同时，提高吞吐量

六、Rejected tasks 拒绝任务
拒绝任务有两种情况：1. 线程池已经被关闭；2. 任务队列已满且maximumPoolSizes已满；
无论哪种情况，都会调用RejectedExecutionHandler的rejectedExecution方法。预定义了四种处理策略：

1. AbortPolicy：默认测策略，抛出RejectedExecutionException运行时异常；
2. CallerRunsPolicy：这提供了一个简单的反馈控制机制，可以减慢提交新任务的速度；
3. DiscardPolicy：直接丢弃新提交的任务；
4. DiscardOldestPolicy：如果执行器没有关闭，队列头的任务将会被丢弃，然后执行器重新尝试执行任务（如果失败，则重复这一过程）；
   我们可以自己定义RejectedExecutionHandler，以适应特殊的容量和队列策略场景中。

七、Hook methods 钩子方法
ThreadPoolExecutor为提供了每个任务执行前后提供了钩子方法，重写beforeExecute(Thread，Runnable)和afterExecute(Runnable，Throwable)方法来操纵执行环境； 例如，重新初始化ThreadLocals，收集统计信息或记录日志等。此外，terminated()在Executor完全终止后需要完成后会被调用，可以重写此方法，以执行任殊处理。
注意：如果hook或回调方法抛出异常，内部的任务线程将会失败并结束。

八、Queue maintenance 维护队列
getQueue()方法可以访问任务队列，一般用于监控和调试。绝不建议将这个方法用于其他目的。当在大量的队列任务被取消时，remove()和purge()方法可用于回收空间。

九、Finalization 关闭

如果程序中不在持有线程池的引用，并且线程池中没有线程时，线程池将会自动关闭。如果您希望确保即使用户忘记调用 shutdown()方法也可以回收未引用的线程池，使未使用线程最终死亡。那么必须通过设置适当的 keep-alive times 并设置allowCoreThreadTimeOut(boolean) 或者 使 corePoolSize下限为0 。
一般情况下，线程池启动后建议手动调用shutdown()关闭。

总结，通过解读ThreadPoolExecutor的注释，我们对ThreadPoolExecutor应该有了比较全面的了解，其实现方式，后续章节详解。

多线程系列目录（不断更新中）：
线程启动原理
线程中断机制
多线程实现方式
FutureTask实现原理
线程池之ThreadPoolExecutor概述
线程池之ThreadPoolExecutor使用
线程池之ThreadPoolExecutor状态控制
线程池之ThreadPoolExecutor执行原理
线程池之ScheduledThreadPoolExecutor概述
线程池的优雅关闭实践

英文原文如下：

/**
 * An {@link ExecutorService} that executes each submitted task using
 * one of possibly several pooled threads, normally configured
 * using {@link Executors} factory methods.
 *
 * <p>Thread pools address two different problems: they usually
 * provide improved performance when executing large numbers of
 * asynchronous tasks, due to reduced per-task invocation overhead,
 * and they provide a means of bounding and managing the resources,
 * including threads, consumed when executing a collection of tasks.
 * Each {@code ThreadPoolExecutor} also maintains some basic
 * statistics, such as the number of completed tasks.
 *
 * <p>To be useful across a wide range of contexts, this class
 * provides many adjustable parameters and extensibility
 * hooks. However, programmers are urged to use the more convenient
 * {@link Executors} factory methods {@link
 * Executors#newCachedThreadPool} (unbounded thread pool, with
 * automatic thread reclamation), {@link Executors#newFixedThreadPool}
 * (fixed size thread pool) and {@link
 * Executors#newSingleThreadExecutor} (single background thread), that
 * preconfigure settings for theost common usage
 * scenarios. Otherwise, use the following guide when manually
 * configuring and tuning this class:
 *
 * <dl>
 *
 * <dt>Core and maximum pool sizes</dt>
 *
 * <dd>A {@code ThreadPoolExecutor} will automatically adjust the
 * pool size (see {@link #getPoolSize})
 * according to the bounds set by
 * corePoolSize (see {@link #getCorePoolSize}) and
 * maximumPoolSize (see {@link #getMaximumPoolSize}).
 *
 * When a new task is submitted in method {@link #execute(Runnable)},
 * and fewer than corePoolSize threads are running, a new thread is
 * created to handle the request, even if other worker threads are
 * idle.  If there are more than corePoolSize but less than
 * maximumPoolSize threads running, a new thread will be created only
 * if the queue is full.  By setting corePoolSize and maximumPoolSize
 * the same, you create a fixed-size thread pool. By setting
 * maximumPoolSize to an essentially unbounded value such as {@code
 * Integer.MAX_VALUE}, you allow the pool to accommodate an arbitrary
 * number of concurrent tasks. Most typically, core and maximum pool
 * sizes are set only upon construction, but they may also be changed
 * dynamically using {@link #setCorePoolSize} and {@link
 * #setMaximumPoolSize}. </dd>
 *
 * <dt>On-demand construction</dt>
 *
 * <dd>By default, even core threads are initially created and
 * started only when new tasks arrive, but this can be overridden
 * dynamically using method {@link #prestartCoreThread} or {@link
 * #prestartAllCoreThreads}.  You probably want to prestart threads if
 * you construct the pool with a non-empty queue. </dd>
 *
 * <dt>Creating new threads</dt>
 *
 * <dd>New threads are created using a {@link ThreadFactory}.  If not
 * otherwise specified, a {@link Executors#defaultThreadFactory} is
 * used, that creates threads to all be in the same {@link
 * ThreadGroup} and with the same {@code NORM_PRIORITY} priority and
 * non-daemon status. By supplying a different ThreadFactory, you can
 * alter the thread's name, thread group, priority, daemon status,
 * etc. If a {@code ThreadFactory} fails to create a thread when asked
 * by returning null from {@code newThread}, the executor will
 * continue, but might not be able to execute any tasks. Threads
 * should possess the "modifyThread" {@code RuntimePermission}. If
 * worker threads or other threads using the pool do not possess this
 * permission, service may be degraded: configuration changes may not
 * take effect in a timely manner, and a shutdown pool may remain in a
 * state in which termination is possible but not completed.</dd>
 *
 * <dt>Keep-alive times</dt>
 *
 * <dd>If the pool currently has more than corePoolSize threads,
 * excess threads will be terminated if they have been idle for more
 * than the keepAliveTime (see {@link #getKeepAliveTime(TimeUnit)}).
 * This provides a means of reducing resource consumption when the
 * pool is not being actively used. If the pool becomes more active
 * later, new threads will be constructed. This parameter can also be
 * changed dynamically using method {@link #setKeepAliveTime(long,
 * TimeUnit)}.  Using a value of {@code Long.MAX_VALUE} {@link
 * TimeUnit#NANOSECONDS} effectively disables idle threads from ever
 * terminating prior to shut down. By default, the keep-alive policy
 * applies only when there are more than corePoolSize threads. But
 * method {@link #allowCoreThreadTimeOut(boolean)} can be used to
 * apply this time-out policy to core threads as well, so long as the
 * keepAliveTime value is non-zero. </dd>
 *
 * <dt>Queuing</dt>
 *
 * <dd>Any {@link BlockingQueue} may be used to transfer and hold
 * submitted tasks.  The use of this queue interacts with pool sizing:
 *
 * <ul>
 *
 * <li> If fewer than corePoolSize threads are running, the Executor
 * always prefers adding a new thread
 * rather than queuing.</li>
 *
 * <li> If corePoolSize or more threads are running, the Executor
 * always prefers queuing a request rather than adding a new
 * thread.</li>
 *
 * <li> If a request cannot be queued, a new thread is created unless
 * this would exceed maximumPoolSize, in which case, the task will be
 * rejected.</li>
 *
 * </ul>
 *
 * There are three general strategies for queuing:
 * <ol>
 *
 * <li> <em> Direct handoffs.</em> A good default choice for a work
 * queue is a {@link SynchronousQueue} that hands off tasks to threads
 * without otherwise holding them. Here, an attempt to queue a task
 * will fail if no threads are immediately available to run it, so a
 * new thread will be constructed. This policy avoids lockups when
 * handling sets of requests that might have internal dependencies.
 * Direct handoffs generally require unbounded maximumPoolSizes to
 * avoid rejection of new submitted tasks. This in turn admits the
 * possibility of unbounded thread growth when commands continue to
 * arrive on average faster than they can be processed.  </li>
 *
 * <li><em> Unbounded queues.</em> Using an unbounded queue (for
 * example a {@link LinkedBlockingQueue} without a predefined
 * capacity) will cause new tasks to wait in the queue when all
 * corePoolSize threads are busy. Thus, no more than corePoolSize
 * threads will ever be created. (And the value of the maximumPoolSize
 * therefore doesn't have any effect.)  This may be appropriate when
 * each task is completely independent of others, so tasks cannot
 * affect each others execution; for example, in a web page server.
 * While this style of queuing can be useful in smoothing out
 * transient bursts of requests, it admits the possibility of
 * unbounded work queue growth when commands continue to arrive on
 * average faster than they can be processed.  </li>
 *
 * <li><em>Bounded queues.</em> A bounded queue (for example, an
 * {@link ArrayBlockingQueue}) helps prevent resource exhaustion when
 * used with finite maximumPoolSizes, but can be more difficult to
 * tune and control.  Queue sizes and maximum pool sizes may be traded
 * off for each other: Using large queues and small pools minimizes
 * CPU usage, OS resources, and context-switching overhead, but can
 * lead to artificially low throughput.  If tasks frequently block (for
 * example if they are I/O bound), a system may be able to schedule
 * time for more threads than you otherwise allow. Use of small queues
 * generally requires larger pool sizes, which keeps CPUs busier but
 * may encounter unacceptable scheduling overhead, which also
 * decreases throughput.  </li>
 *
 * </ol>
 *
 * </dd>
 *
 * <dt>Rejected tasks</dt>
 *
 * <dd>New tasks submitted in method {@link #execute(Runnable)} will be
 * <em>rejected</em> when the Executor has been shut down, and also when
 * the Executor uses finite bounds for both maximum threads and work queue
 * capacity, and is saturated.  In either case, the {@code execute} method
 * invokes the {@link
 * RejectedExecutionHandler#rejectedExecution(Runnable, ThreadPoolExecutor)}
 * method of its {@link RejectedExecutionHandler}.  Four predefined handler
 * policies are provided:
 *
 * <ol>
 *
 * <li> In the default {@link ThreadPoolExecutor.AbortPolicy}, the
 * handler throws a runtime {@link RejectedExecutionException} upon
 * rejection. </li>
 *
 * <li> In {@link ThreadPoolExecutor.CallerRunsPolicy}, the thread
 * that invokes {@code execute} itself runs the task. This provides a
 * simple feedback control mechanism that will slow down the rate that
 * new tasks are submitted. </li>
 *
 * <li> In {@link ThreadPoolExecutor.DiscardPolicy}, a task that
 * cannot be executed is simply dropped.  </li>
 *
 * <li>In {@link ThreadPoolExecutor.DiscardOldestPolicy}, if the
 * executor is not shut down, the task at the head of the work queue
 * is dropped, and then execution is retried (which can fail again,
 * causing this to be repeated.) </li>
 *
 * </ol>
 *
 * It is possible to define and use other kinds of {@link
 * RejectedExecutionHandler} classes. Doing so requires some care
 * especially when policies are designed to work only under particular
 * capacity or queuing policies. </dd>
 *
 * <dt>Hook methods</dt>
 *
 * <dd>This class provides {@code protected} overridable
 * {@link #beforeExecute(Thread, Runnable)} and
 * {@link #afterExecute(Runnable, Throwable)} methods that are called
 * before and after execution of each task.  These can be used to
 * manipulate the execution environment; for example, reinitializing
 * ThreadLocals, gathering statistics, or adding log entries.
 * Additionally, method {@link #terminated} can be overridden to perform
 * any special processing that needs to be done once the Executor has
 * fully terminated.
 *
 * <p>If hook or callback methods throw exceptions, internal worker
 * threads may in turn fail and abruptly terminate.</dd>
 *
 * <dt>Queue maintenance</dt>
 *
 * <dd>Method {@link #getQueue()} allows access to the work queue
 * for purposes of monitoring and debugging.  Use of this method for
 * any other purpose is strongly discouraged.  Two supplied methods,
 * {@link #remove(Runnable)} and {@link #purge} are available to
 * assist in storage reclamation when large numbers of queued tasks
 * become cancelled.</dd>
 *
 * <dt>Finalization</dt>
 *
 * <dd>A pool that is no longer referenced in a program <em>AND</em>
 * has no remaining threads will be {@code shutdown} automatically. If
 * you would like to ensure that unreferenced pools are reclaimed even
 * if users forget to call {@link #shutdown}, then you must arrange
 * that unused threads eventually die, by setting appropriate
 * keep-alive times, using a lower bound of zero core threads and/or
 * setting {@link #allowCoreThreadTimeOut(boolean)}.  </dd>
 *
 * </dl>
 *
 * @since 1.5
 * @author Doug Lea
 */