聊聊如何自定义parallelStream的线程池

序

本文主要研究一下parallelStream怎么使用自定义的线程池

ForkJoinPool

java/util/concurrent/ForkJoinPool.java

public class ForkJoinPool extends AbstractExecutorService {

    public ForkJoinPool(int parallelism,
                        ForkJoinWorkerThreadFactory factory,
                        UncaughtExceptionHandler handler,
                        boolean asyncMode) {
        this(checkParallelism(parallelism),
             checkFactory(factory),
             handler,
             asyncMode ? FIFO_QUEUE : LIFO_QUEUE,
             "ForkJoinPool-" + nextPoolId() + "-worker-");
        checkPermission();
    }

    private ForkJoinPool(int parallelism,
                         ForkJoinWorkerThreadFactory factory,
                         UncaughtExceptionHandler handler,
                         int mode,
                         String workerNamePrefix) {
        this.workerNamePrefix = workerNamePrefix;
        this.factory = factory;
        this.ueh = handler;
        this.config = (parallelism & SMASK) | mode;
        long np = (long)(-parallelism); // offset ctl counts
        this.ctl = ((np << AC_SHIFT) & AC_MASK) | ((np << TC_SHIFT) & TC_MASK);
    }

    private static ForkJoinPool makeCommonPool() {
        int parallelism = -1;
        ForkJoinWorkerThreadFactory factory = null;
        UncaughtExceptionHandler handler = null;
        try {  // ignore exceptions in accessing/parsing properties
            String pp = System.getProperty
                ("java.util.concurrent.ForkJoinPool.common.parallelism");
            String fp = System.getProperty
                ("java.util.concurrent.ForkJoinPool.common.threadFactory");
            String hp = System.getProperty
                ("java.util.concurrent.ForkJoinPool.common.exceptionHandler");
            if (pp != null)
                parallelism = Integer.parseInt(pp);
            if (fp != null)
                factory = ((ForkJoinWorkerThreadFactory)ClassLoader.
                           getSystemClassLoader().loadClass(fp).newInstance());
            if (hp != null)
                handler = ((UncaughtExceptionHandler)ClassLoader.
                           getSystemClassLoader().loadClass(hp).newInstance());
        } catch (Exception ignore) {
        }
        if (factory == null) {
            if (System.getSecurityManager() == null)
                factory = new DefaultCommonPoolForkJoinWorkerThreadFactory();
            else // use security-managed default
                factory = new InnocuousForkJoinWorkerThreadFactory();
        }
        if (parallelism < 0 && // default 1 less than #cores
            (parallelism = Runtime.getRuntime().availableProcessors() - 1) <= 0)
            parallelism = 1;
        if (parallelism > MAX_CAP)
            parallelism = MAX_CAP;
        return new ForkJoinPool(parallelism, factory, handler, LIFO_QUEUE,
                                "ForkJoinPool.commonPool-worker-");
    }

}

parallelStream默认使用的是common的ForkJoinPool，可以通过系统属性来设置parallelism等

ForkJoinPoolFactoryBean

org/springframework/scheduling/concurrent/ForkJoinPoolFactoryBean.java

public class ForkJoinPoolFactoryBean implements FactoryBean, InitializingBean, DisposableBean {

	private boolean commonPool = false;

	private int parallelism = Runtime.getRuntime().availableProcessors();

	private ForkJoinPool.ForkJoinWorkerThreadFactory threadFactory = ForkJoinPool.defaultForkJoinWorkerThreadFactory;

	@Nullable
	private Thread.UncaughtExceptionHandler uncaughtExceptionHandler;

	private boolean asyncMode = false;

	private int awaitTerminationSeconds = 0;

	@Nullable
	private ForkJoinPool forkJoinPool;

	//......

	@Override
	public void destroy() {
		if (this.forkJoinPool != null) {
			// Ignored for the common pool.
			this.forkJoinPool.shutdown();

			// Wait for all tasks to terminate - works for the common pool as well.
			if (this.awaitTerminationSeconds > 0) {
				try {
					this.forkJoinPool.awaitTermination(this.awaitTerminationSeconds, TimeUnit.SECONDS);
				}
				catch (InterruptedException ex) {
					Thread.currentThread().interrupt();
				}
			}
		}
	}

}	

spring3.1提供了ForkJoinPoolFactoryBean，可以用于创建并托管forkJoinPool

示例

配置

@Configuration
public class ForkJoinConfig {

    @Bean
    public ForkJoinPoolFactoryBean forkJoinPoolFactoryBean() {
        ForkJoinPoolFactoryBean factoryBean = new ForkJoinPoolFactoryBean();
        factoryBean.setCommonPool(false);
        // NOTE LIFO_QUEUE FOR working steal from tail of queue
        factoryBean.setAsyncMode(true); // NOTE true FIFO_QUEUE, false LIFO_QUEUE
        factoryBean.setParallelism(10);
        // factoryBean.setUncaughtExceptionHandler();
        factoryBean.setAwaitTerminationSeconds(60);
        return factoryBean;
    }
}

使用

    @Autowired
    ForkJoinPoolFactoryBean forkJoinPoolFactoryBean;

    public void streamParallel() throws ExecutionException, InterruptedException {
        List result = forkJoinPoolFactoryBean.getObject().submit(new Callable>() {
            @Override
            public List call() throws Exception {
                return IntStream.rangeClosed(1, 20).parallel().mapToObj(i -> {
                    log.info("thread:{}", Thread.currentThread().getName());
                    return new TodoTask(i, "name"+i);
                }).collect(Collectors.toList());
            }
        }).get();
        result.stream().forEach(System.out::println);
    }

common的workerName前缀为ForkJoinPool.commonPool-worker-
自定义的workerName前缀默认为ForkJoinPool- nextPoolId() -worker-

小结

parallelStream默认使用的是commonPool，是static代码块默认初始化，针对个别场景可以自定义ForkJoinPool，将parallelStream作为一个任务丢进去，这样子就不会影响默认的commonPool。