Future.get() 在配置RejectedExecutionHandler为ThreadPoolExecutor.DiscardPolicy策略时一直阻塞

转载自：https://blog.youkuaiyun.com/doctor_who2004/article/details/115647638

1、先复现这种情况：

package com.sdcuike.java11;

import java.util.concurrent.*;

import com.google.common.util.concurrent.ThreadFactoryBuilder;

public class Demo {

    public static void main(String[] args) throws ExecutionException, InterruptedException {
        ThreadFactory threadFactory = new ThreadFactoryBuilder().setNameFormat("rpc-pool-%d").setDaemon(true).build();
        ThreadPoolExecutor threadPoolExecutor = new ThreadPoolExecutor(1, 1, 1, TimeUnit.HOURS, new SynchronousQueue<>(), threadFactory, new ThreadPoolExecutor.DiscardPolicy() {
            @Override
            public void rejectedExecution(Runnable r, ThreadPoolExecutor executor) {
                System.out.println("rejectedExecution");
     
                super.rejectedExecution(r, executor);
            }
        });
     
        threadPoolExecutor.submit(new Runnable() {
            @Override
            public void run() {
                try {
                    TimeUnit.HOURS.sleep(1);
                } catch (InterruptedException e) {
                    e.printStackTrace();
                }
            }
        });
     
        Future<String> future = threadPoolExecutor.submit(new Callable<String>() {
     
            @Override
            public String call() throws Exception {
                return "done";
            }
        });
     
        String result = future.get();
        System.out.println(result);
        System.out.println("done....");
    }

}

线程池中线程的配置为daemon线程（后台运行），当我们的main线程退出时，后台线程也会退出。会输出

done… 而现实的结果是main线程一直阻塞在future.get()调用：

2、why？

我们看下源码：java.util.concurrent.FutureTask#get()

public V get() throws InterruptedException, ExecutionException {
        int s = state;
        if (s <= COMPLETING)
            s = awaitDone(false, 0L);
        return report(s);
    }

FutureTask 内部有几种状态：

private static final int NEW          = 0;
    private static final int COMPLETING   = 1;
    private static final int NORMAL       = 2;
    private static final int EXCEPTIONAL  = 3;
    private static final int CANCELLED    = 4;
    private static final int INTERRUPTING = 5;
    private static final int INTERRUPTED  = 6;

当状态state<=COMPLETING 即COMPLETING和NEW两种状态时，会一直阻塞：awaitDone(false, 0L)。

FutureTask的状态设置是在线程池ThreadPoolExecutor执行过程中所设置的，一般情况下线程池队列不满，即不会执行RejectedExecutionHandler处理不能处理的任务时，状态都会设置为state > COMPLETING的某个状态 ，但线程池执行RejectedExecutionHandler的时候，java内置的拒绝策略有：

CallerRunsPolicy
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();
            }
        }
    }

这种策略会使用调用线程池执行任务的当前线程执行，java.util.concurrent.FutureTask#run() 执行肯定会设置新的完成状态。

AbortPolicy
public static class AbortPolicy implements RejectedExecutionHandler {
        /**
         * Creates an {@code AbortPolicy}.
         */
        public AbortPolicy() { }

        /**
         * Always throws RejectedExecutionException.
         *
         * @param r the runnable task requested to be executed
         * @param e the executor attempting to execute this task
         * @throws RejectedExecutionException always
         */
        public void rejectedExecution(Runnable r, ThreadPoolExecutor e) {
            throw new RejectedExecutionException("Task " + r.toString() +
                                                 " rejected from " +
                                                 e.toString());
        }

这种策略是线程池没配置的情况下使用的 默认策略，直接抛出异常，线程池根本就不会执行任务。

DiscardOldestPolicy
public static class DiscardOldestPolicy implements RejectedExecutionHandler {
        /**
         * Creates a {@code DiscardOldestPolicy} for the given executor.
         */
        public DiscardOldestPolicy() { }

        /**
         * Obtains and ignores the next task that the executor
         * would otherwise execute, if one is immediately available,
         * and then retries execution of task r, unless the executor
         * is shut down, in which case task r is instead 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()) {
                e.getQueue().poll();
                e.execute(r);
            }
        }
    }

丢弃任务队列队头的任务，任何重试执行提交的新任务。状态肯定会设置。

DiscardPolicy
public static class DiscardPolicy implements RejectedExecutionHandler {
        /**
         * Creates a {@code DiscardPolicy}.
         */
        public DiscardPolicy() { }

        /**
         * Does nothing, which has the effect of discarding task r.
         *
         * @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) {
        }
    }

这种策略什么都不做，任务的状态自然是初始状态，永远不会更新任务的状态，导致java.util.concurrent.FutureTask#get() 一直阻塞。

所以，最好的情况下是不使用这种策略，或者使用java.util.concurrent.FutureTask#get(long, java.util.concurrent.TimeUnit) 带有超时的方法获取结果，或者重写策略，取消任务： ((FutureTask) r).cancel(true);

 @Override
            public void rejectedExecution(Runnable r, ThreadPoolExecutor executor) {
                System.out.println("rejectedExecution");
                ((FutureTask) r).cancel(true)
                super.rejectedExecution(r, executor);
            }

3、解决一直阻塞的问题总结

1、重写DiscardPolicy，取消任务： ((FutureTask) r).cancel(true)。
2、不使用DiscardPolicy，建议还是用线程池默认的AbortPolicy，可以重写，打印日志，再调用super.rejectedExecution(r, executor)，抛出默认异常行为。
3、使用java.util.concurrent.FutureTask#get(long, java.util.concurrent.TimeUnit) 带有超时的方法获取结果