Java并发 -- Worker Thread模式

Worker Thread模式

Worker Thread模式可以类比现实世界里车间的工作模式，Worker Thread对应车间里的工人（人数确定）
用阻塞队列做任务池，然后创建固定数量的线程消费阻塞队列中的任务 – 这就是Java中的线程池方案

echo服务

private ExecutorService pool = Executors.newFixedThreadPool(500);

public void handle() throws IOException {
    // 处理请求
    try (ServerSocketChannel ssc = ServerSocketChannel.open().bind(new InetSocketAddress(8080))) {
        while (true) {
            // 接收请求
            SocketChannel sc = ssc.accept();
            // 将请求处理任务提交给线程池
            pool.execute(() -> {
                try {
                    // 读Socket
                    ByteBuffer rb = ByteBuffer.allocateDirect(1024);
                    sc.read(rb);
                    TimeUnit.SECONDS.sleep(1);
                    // 写Socket
                    ByteBuffer wb = (ByteBuffer) rb.flip();
                    sc.write(wb);
                    sc.close();
                } catch (IOException | InterruptedException ignored) {
                }
            });
        }
    } finally {
        pool.shutdown();
    }
}

正确地创建线程池

Java线程池既能避免无限制地创建线程导致OOM，也能避免无限制地接收任务导致OOM（有界队列）
当请求量大于有界队列的容量时，应该合理地拒绝请求，在创建线程池时，应该清晰地指明拒绝策略
为了便于调试和诊断问题，在实际工作中应该给线程赋予一个业务相关的命名

private ExecutorService pool = new ThreadPoolExecutor(50, 500, 60L, TimeUnit.SECONDS,
        new LinkedBlockingQueue<>(2000), // 有界队列
        runnable -> new Thread(runnable, "echo-" + runnable.hashCode()), // ThreadFactory
        new ThreadPoolExecutor.CallerRunsPolicy()); // 拒绝策略

避免线程死锁

如果提交到相同线程池的任务不是相互独立的，而是有依赖关系的，有可能会导致线程死锁
通用解决方案：为不同的任务创建不同的线程池，提交到相同线程池中的任务一定是相互独立的
下图中第一阶段的任务会等待第二阶段的子任务完成

// L1、L2阶段共用线程池
ExecutorService pool = Executors.newFixedThreadPool(2);
CountDownLatch l1 = new CountDownLatch(2);
for (int i = 0; i < 2; i++) {
    System.out.println("L1");
    pool.execute(() -> {
        CountDownLatch l2 = new CountDownLatch(2);
        for (int j = 0; j < 2; j++) {
            pool.execute(() -> {
                System.out.println("L2");
                l2.countDown();
            });
        }
        try {
            // 线程池中的2个线程都阻塞在l2.await()，没有多余线程去执行L2阶段的任务（在线程池的任务队列中等待）
            l2.await(); // line 28
        } catch (InterruptedException ignored) {
        }
        l1.countDown();
    });
}
l1.await();
// 输出
//  L1
//  L1

jstack

// 阻塞在l2.await()
"pool-1-thread-2" #11 prio=5 os_prio=31 tid=0x00007f934e8f5000 nid=0x4303 waiting on condition [0x000070000792d000]
   java.lang.Thread.State: WAITING (parking)
    at sun.misc.Unsafe.park(Native Method)
    - parking to wait for  <0x000000079609bd58> (a java.util.concurrent.CountDownLatch$Sync)
    at java.util.concurrent.locks.LockSupport.park(LockSupport.java:175)
    at java.util.concurrent.locks.AbstractQueuedSynchronizer.parkAndCheckInterrupt(AbstractQueuedSynchronizer.java:836)
    at java.util.concurrent.locks.AbstractQueuedSynchronizer.doAcquireSharedInterruptibly(AbstractQueuedSynchronizer.java:997)
    at java.util.concurrent.locks.AbstractQueuedSynchronizer.acquireSharedInterruptibly(AbstractQueuedSynchronizer.java:1304)
    at java.util.concurrent.CountDownLatch.await(CountDownLatch.java:231)
    at time.geek.worker.thread.DeadLock.lambda$deadLockTest$1(DeadLock.java:28)
    at time.geek.worker.thread.DeadLock$$Lambda$1/1221555852.run(Unknown Source)
    at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
    at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
    at java.lang.Thread.run(Thread.java:748)

// 阻塞在l2.await()
"pool-1-thread-1" #10 prio=5 os_prio=31 tid=0x00007f9350142800 nid=0x3c03 waiting on condition [0x000070000782a000]
   java.lang.Thread.State: WAITING (parking)
    at sun.misc.Unsafe.park(Native Method)
    - parking to wait for  <0x0000000795ff56a8> (a java.util.concurrent.CountDownLatch$Sync)
    at java.util.concurrent.locks.LockSupport.park(LockSupport.java:175)
    at java.util.concurrent.locks.AbstractQueuedSynchronizer.parkAndCheckInterrupt(AbstractQueuedSynchronizer.java:836)
    at java.util.concurrent.locks.AbstractQueuedSynchronizer.doAcquireSharedInterruptibly(AbstractQueuedSynchronizer.java:997)
    at java.util.concurrent.locks.AbstractQueuedSynchronizer.acquireSharedInterruptibly(AbstractQueuedSynchronizer.java:1304)
    at java.util.concurrent.CountDownLatch.await(CountDownLatch.java:231)
    at time.geek.worker.thread.DeadLock.lambda$deadLockTest$1(DeadLock.java:28)
    at time.geek.worker.thread.DeadLock$$Lambda$1/1221555852.run(Unknown Source)
    at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
    at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
    at java.lang.Thread.run(Thread.java:748)

对比Thread-Per-Message模式

Thread-Per-Message模式：主线程直接创建子线程，主子线程之间可以直接通信
Worker Thread模式：主线程提交任务到线程池，但主线程并不关心任务被哪个线程执行
- 能够避免线程频繁创建、销毁的问题，并且能够限制线程的最大数量
- Java利用Worker Thread模式来实现线程池