mirror of
https://github.com/LingCoder/OnJava8.git
synced 2026-08-24 06:53:27 +08:00
Merge pull request #171 from blackwatchcup/master
Creating and Running Tasks
This commit is contained in:
@@ -711,10 +711,487 @@ public class ParallelStreamPuzzle3 {
|
||||
- 并行流只看起来很容易
|
||||
|
||||
实际上,在许多情况下,并行流确实可以毫不费力地更快地产生结果。但正如您所见,只需将**parallel()**打到您的Stream操作上并不一定是安全的事情。在使用**parallel()**之前,您必须了解并行性如何帮助或损害您的操作。有个错误认识是认为并行性总是一个好主意。事实上并不是。Stream意味着您不需要重写所有代码以便并行运行它。流什么都不做的是取代理解并行性如何工作的需要,以及它是否有助于实现您的目标。
|
||||
|
||||
<!-- Creating and Running Tasks -->
|
||||
## 创建和运行任务
|
||||
|
||||
如果无法通过并行流实现并发,则必须创建并运行自己的任务。稍后您将看到运行任务的理想Java 8方法是CompletableFuture,但我们将使用更基本的工具介绍概念。
|
||||
|
||||
Java并发的历史始于非常原始和有问题的机制,并且充满了各种尝试的改进。这些主要归入附录:[低级并发(Appendix: Low-Level Concurrency)](./Appendix-Low-Level-Concurrency.md)。在这里,我们将展示一个规范形式,表示创建和运行任务的最简单,最好的方法。与并发中的所有内容一样,存在各种变体,但这些变体要么降级到该附录,要么超出本书的范围。
|
||||
|
||||
- Tasks and Executors
|
||||
|
||||
在Java的早期版本中,您通过直接创建自己的Thread对象来使用线程,甚至将它们子类化以创建您自己的特定“任务线程”对象。你手动调用了构造函数并自己启动了线程。
|
||||
|
||||
创建所有这些线程的开销变得非常重要,现在不鼓励采用实际操作方法。在Java 5中,添加了类来为您处理线程池。您可以将任务创建为单独的类型,然后将其交给ExecutorService以运行该任务,而不是为每种不同类型的任务创建新的Thread子类型。ExecutorService为您管理线程,并且在运行任务后重新循环线程而不是丢弃线程。
|
||||
|
||||
首先,我们将创建一个几乎不执行任务的任务。它“sleep”(暂停执行)100毫秒,显示其标识符和正在执行任务的线程的名称,然后完成:
|
||||
|
||||
```java
|
||||
// concurrent/NapTask.java
|
||||
import onjava.Nap;
|
||||
public class NapTask implements Runnable {
|
||||
finalint id;
|
||||
public NapTask(int id) {
|
||||
this.id = id;
|
||||
}
|
||||
@Override
|
||||
public void run() {
|
||||
new Nap(0.1);// Seconds
|
||||
System.out.println(this + " "+
|
||||
Thread.currentThread().getName());
|
||||
}
|
||||
@Override
|
||||
public String toString() {
|
||||
return"NapTask[" + id + "]";
|
||||
}
|
||||
}
|
||||
```
|
||||
|
||||
这只是一个**Runnable**:一个包含**run()**方法的类。它没有包含实际运行任务的机制。我们使用**Nap**类中的“sleep”:
|
||||
|
||||
```java
|
||||
// onjava/Nap.java
|
||||
package onjava;
|
||||
import java.util.concurrent.*;
|
||||
public class Nap {
|
||||
public Nap(double t) { // Seconds
|
||||
try {
|
||||
TimeUnit.MILLISECONDS.sleep((int)(1000 * t));
|
||||
} catch(InterruptedException e){
|
||||
throw new RuntimeException(e);
|
||||
}
|
||||
}
|
||||
public Nap(double t, String msg) {
|
||||
this(t);
|
||||
System.out.println(msg);
|
||||
}
|
||||
}
|
||||
```
|
||||
为了消除异常处理的视觉噪声,这被定义为实用程序。第二个构造函数在超时时显示一条消息
|
||||
|
||||
对**TimeUnit.MILLISECONDS.sleep()**的调用获取“当前线程”并在参数中将其置于休眠状态,这意味着该线程被挂起。这并不意味着底层处理器停止。操作系统将其切换到其他任务,例如在您的计算机上运行另一个窗口。OS任务管理器定期检查**sleep()**是否超时。当它执行时,线程被“唤醒”并给予更多处理时间。
|
||||
|
||||
你可以看到**sleep()**抛出一个已检查的**InterruptedException**;这是原始Java设计中的一个工件,它通过突然断开它们来终止任务。因为它往往会产生不稳定的状态,所以后来不鼓励终止。但是,我们必须在需要或仍然发生终止的情况下捕获异常。
|
||||
|
||||
要执行任务,我们将从最简单的方法--SingleThreadExecutor开始:
|
||||
|
||||
```java
|
||||
/ concurrent/SingleThreadExecutor.java
|
||||
import java.util.concurrent.*;
|
||||
import java.util.stream.*;
|
||||
import onjava.*;
|
||||
public class SingleThreadExecutor {
|
||||
public static void main(String[] args) {
|
||||
ExecutorService exec =
|
||||
Executors.newSingleThreadExecutor();
|
||||
IntStream.range(0, 10)
|
||||
.mapToObj(NapTask::new)
|
||||
.forEach(exec::execute);
|
||||
System.out.println("All tasks submitted");
|
||||
exec.shutdown();
|
||||
while(!exec.isTerminated()) {
|
||||
System.out.println(
|
||||
Thread.currentThread().getName()+
|
||||
" awaiting termination");
|
||||
new Nap(0.1);
|
||||
}
|
||||
}
|
||||
}
|
||||
/* Output:
|
||||
All tasks submitted
|
||||
main awaiting termination
|
||||
main awaiting termination
|
||||
NapTask[0] pool-1-thread-1
|
||||
main awaiting termination
|
||||
NapTask[1] pool-1-thread-1
|
||||
main awaiting termination
|
||||
NapTask[2] pool-1-thread-1
|
||||
main awaiting termination
|
||||
NapTask[3] pool-1-thread-1
|
||||
main awaiting termination
|
||||
NapTask[4] pool-1-thread-1
|
||||
main awaiting termination
|
||||
NapTask[5] pool-1-thread-1
|
||||
main awaiting termination
|
||||
NapTask[6] pool-1-thread-1
|
||||
main awaiting termination
|
||||
NapTask[7] pool-1-thread-1
|
||||
main awaiting termination
|
||||
NapTask[8] pool-1-thread-1
|
||||
main awaiting termination
|
||||
NapTask[9] pool-1-thread-1
|
||||
*/
|
||||
```
|
||||
|
||||
首先请注意,没有**SingleThreadExecutor**类。**newSingleThreadExecutor()**是**Executors**中的工厂,它创建特定类型的[^4]
|
||||
|
||||
我创建了十个NapTasks并将它们提交给ExecutorService,这意味着它们开始自己运行。然而,在此期间,main()继续做事。当我运行callexec.shutdown()时,它告诉ExecutorService完成已经提交的任务,但不接受任何新任务。此时,这些任务仍然在运行,因此我们必须等到它们在退出main()之前完成。这是通过检查exec.isTerminated()来实现的,这在所有任务完成后变为true。
|
||||
|
||||
请注意,main()中线程的名称是main,并且只有一个其他线程pool-1-thread-1。此外,交错输出显示两个线程确实同时运行。
|
||||
|
||||
如果你只是调用exec.shutdown(),程序将完成所有任务。也就是说,虽然不需要(!exec.isTerminated())。
|
||||
|
||||
```java
|
||||
// concurrent/SingleThreadExecutor2.java
|
||||
import java.util.concurrent.*;
|
||||
import java.util.stream.*;
|
||||
public class SingleThreadExecutor2 {
|
||||
public static void main(String[] args)throws InterruptedException {
|
||||
ExecutorService exec
|
||||
=Executors.newSingleThreadExecutor();
|
||||
IntStream.range(0, 10)
|
||||
.mapToObj(NapTask::new)
|
||||
.forEach(exec::execute);
|
||||
exec.shutdown();
|
||||
}
|
||||
}
|
||||
/* Output:
|
||||
NapTask[0] pool-1-thread-1
|
||||
NapTask[1] pool-1-thread-1
|
||||
NapTask[2] pool-1-thread-1
|
||||
NapTask[3] pool-1-thread-1
|
||||
NapTask[4] pool-1-thread-1
|
||||
NapTask[5] pool-1-thread-1
|
||||
NapTask[6] pool-1-thread-1
|
||||
NapTask[7] pool-1-thread-1
|
||||
NapTask[8] pool-1-thread-1
|
||||
NapTask[9] pool-1-thread-1
|
||||
*/
|
||||
```
|
||||
|
||||
一旦你callexec.shutdown(),尝试提交新任务将抛出RejectedExecutionException。
|
||||
|
||||
```java
|
||||
// concurrent/MoreTasksAfterShutdown.java
|
||||
import java.util.concurrent.*;
|
||||
public class MoreTasksAfterShutdown {
|
||||
public static void main(String[] args) {
|
||||
ExecutorService exec
|
||||
=Executors.newSingleThreadExecutor();
|
||||
exec.execute(newNapTask(1));
|
||||
exec.shutdown();
|
||||
try {
|
||||
exec.execute(newNapTask(99));
|
||||
} catch(RejectedExecutionException e) {
|
||||
System.out.println(e);
|
||||
}
|
||||
}
|
||||
}
|
||||
/* Output:
|
||||
java.util.concurrent.RejectedExecutionException: TaskNapTask[99] rejected from java.util.concurrent.ThreadPoolExecutor@4e25154f[Shutting down, pool size = 1,active threads = 1, queued tasks = 0, completed tasks =0]NapTask[1] pool-1-thread-1
|
||||
*/
|
||||
|
||||
```
|
||||
|
||||
**exec.shutdown()**的替代方法是**exec.shutdownNow()**,它除了不接受新任务外,还会尝试通过中断任务来停止任何当前正在运行的任务。同样,中断是错误的,容易出错并且不鼓励。
|
||||
|
||||
- 使用更多线程
|
||||
|
||||
使用线程的重点是(几乎总是)更快地完成任务,那么我们为什么要限制自己使用SingleThreadExecutor呢?查看执行**Executors**的Javadoc,您将看到更多选项。例如CachedThreadPool:
|
||||
|
||||
```java
|
||||
// concurrent/CachedThreadPool.java
|
||||
import java.util.concurrent.*;
|
||||
import java.util.stream.*;
|
||||
public class CachedThreadPool {
|
||||
public static void main(String[] args) {
|
||||
ExecutorService exec
|
||||
=Executors.newCachedThreadPool();
|
||||
IntStream.range(0, 10)
|
||||
.mapToObj(NapTask::new)
|
||||
.forEach(exec::execute);
|
||||
exec.shutdown();
|
||||
}
|
||||
}
|
||||
/* Output:
|
||||
NapTask[7] pool-1-thread-8
|
||||
NapTask[4] pool-1-thread-5
|
||||
NapTask[1] pool-1-thread-2
|
||||
NapTask[3] pool-1-thread-4
|
||||
NapTask[0] pool-1-thread-1
|
||||
NapTask[8] pool-1-thread-9
|
||||
NapTask[2] pool-1-thread-3
|
||||
NapTask[9] pool-1-thread-10
|
||||
NapTask[6] pool-1-thread-7
|
||||
NapTask[5] pool-1-thread-6
|
||||
*/
|
||||
|
||||
```
|
||||
|
||||
当你运行这个程序时,你会发现它完成得更快。这是有道理的,而不是使用相同的线程来顺序运行每个任务,每个任务都有自己的线程,所以它们都并行运行。似乎没有缺点,很难看出为什么有人会使用SingleThreadExecutor。
|
||||
|
||||
要理解这个问题,我们需要一个更复杂的任务:
|
||||
|
||||
```java
|
||||
// concurrent/InterferingTask.java
|
||||
public class InterferingTask implements Runnable {
|
||||
final int id;
|
||||
private static Integer val = 0;
|
||||
public InterferingTask(int id) {
|
||||
this.id = id;
|
||||
}
|
||||
@Override
|
||||
public void run() {
|
||||
for(int i = 0; i < 100; i++)
|
||||
val++;
|
||||
System.out.println(id + " "+
|
||||
Thread.currentThread().getName() + " " + val);
|
||||
}
|
||||
}
|
||||
|
||||
```
|
||||
|
||||
每个任务增加val一百次。这似乎很简单。让我们用CachedThreadPool尝试一下:
|
||||
|
||||
```java
|
||||
// concurrent/CachedThreadPool2.java
|
||||
import java.util.concurrent.*;
|
||||
import java.util.stream.*;
|
||||
public class CachedThreadPool2 {
|
||||
public static void main(String[] args) {
|
||||
ExecutorService exec
|
||||
=Executors.newCachedThreadPool();
|
||||
IntStream.range(0, 10)
|
||||
.mapToObj(InterferingTask::new)
|
||||
.forEach(exec::execute);
|
||||
exec.shutdown();
|
||||
}
|
||||
}
|
||||
/* Output:
|
||||
0 pool-1-thread-1 200
|
||||
1 pool-1-thread-2 200
|
||||
4 pool-1-thread-5 300
|
||||
5 pool-1-thread-6 400
|
||||
8 pool-1-thread-9 500
|
||||
9 pool-1-thread-10 600
|
||||
2 pool-1-thread-3 700
|
||||
7 pool-1-thread-8 800
|
||||
3 pool-1-thread-4 900
|
||||
6 pool-1-thread-7 1000
|
||||
*/
|
||||
|
||||
```
|
||||
|
||||
输出不是我们所期望的,并且从一次运行到下一次运行会有所不同。问题是所有的任务都试图写入val的单个实例,并且他们正在踩着彼此的脚趾。我们说这样的类不是线程安全的。让我们看看SingleThreadExecutor会发生什么:
|
||||
|
||||
```java
|
||||
// concurrent/SingleThreadExecutor3.java
|
||||
import java.util.concurrent.*;
|
||||
import java.util.stream.*;
|
||||
public class SingleThreadExecutor3 {
|
||||
public static void main(String[] args)throws InterruptedException {
|
||||
ExecutorService exec
|
||||
=Executors.newSingleThreadExecutor();
|
||||
IntStream.range(0, 10)
|
||||
.mapToObj(InterferingTask::new)
|
||||
.forEach(exec::execute);
|
||||
exec.shutdown();
|
||||
}
|
||||
}
|
||||
/* Output:
|
||||
0 pool-1-thread-1 100
|
||||
1 pool-1-thread-1 200
|
||||
2 pool-1-thread-1 300
|
||||
3 pool-1-thread-1 400
|
||||
4 pool-1-thread-1 500
|
||||
5 pool-1-thread-1 600
|
||||
6 pool-1-thread-1 700
|
||||
7 pool-1-thread-1 800
|
||||
8 pool-1-thread-1 900
|
||||
9 pool-1-thread-1 1000
|
||||
*/
|
||||
|
||||
```
|
||||
|
||||
现在我们每次都得到一致的结果,尽管**InterferingTask**缺乏线程安全性。这是SingleThreadExecutor的主要好处 - 因为它一次运行一个任务,这些任务不会相互干扰,因此强加了线程安全性。这种现象称为线程限制,因为在单线程上运行任务限制了它们的影响。线程限制限制了加速,但可以节省很多困难的调试和重写。
|
||||
|
||||
- 产生结果
|
||||
|
||||
因为**InterferingTask**是一个**Runnable**,它没有返回值,因此只能使用副作用产生结果 - 操纵缓冲值而不是返回结果。副作用是并发编程中的主要问题之一,因为我们看到了**CachedThreadPool2.java**。**InterferingTask**中的**val**被称为可变共享状态,这就是问题所在:多个任务同时修改同一个变量会产生竞争。结果取决于首先在终点线上执行哪个任务,并修改变量(以及其他可能性的各种变化)。
|
||||
|
||||
避免竞争条件的最好方法是避免可变的共享状态。我们可以称之为自私的孩子原则:什么都不分享。
|
||||
|
||||
使用**InterferingTask**,最好删除副作用并返回任务结果。为此,我们创建**Callable**而不是**Runnable**:
|
||||
|
||||
```java
|
||||
// concurrent/CountingTask.java
|
||||
import java.util.concurrent.*;
|
||||
public class CountingTask implements Callable<Integer> {
|
||||
final int id;
|
||||
public CountingTask(int id) { this.id = id; }
|
||||
@Override
|
||||
public Integer call() {
|
||||
Integer val = 0;
|
||||
for(int i = 0; i < 100; i++)
|
||||
val++;
|
||||
System.out.println(id + " " +
|
||||
Thread.currentThread().getName() + " " + val);
|
||||
return val;
|
||||
}
|
||||
}
|
||||
|
||||
```
|
||||
|
||||
**call()完全独立于所有其他CountingTasks生成其结果**,这意味着没有可变的共享状态
|
||||
|
||||
**ExecutorService**允许您使用**invokeAll()**启动集合中的每个Callable:
|
||||
|
||||
```java
|
||||
// concurrent/CachedThreadPool3.java
|
||||
import java.util.*;
|
||||
import java.util.concurrent.*;
|
||||
import java.util.stream.*;
|
||||
public class CachedThreadPool3 {
|
||||
public static Integer extractResult(Future<Integer> f) {
|
||||
try {
|
||||
return f.get();
|
||||
} catch(Exception e) {
|
||||
throw new RuntimeException(e);
|
||||
}
|
||||
}
|
||||
public static void main(String[] args)throws InterruptedException {
|
||||
ExecutorService exec =
|
||||
Executors.newCachedThreadPool();
|
||||
List<CountingTask> tasks =
|
||||
IntStream.range(0, 10)
|
||||
.mapToObj(CountingTask::new)
|
||||
.collect(Collectors.toList());
|
||||
List<Future<Integer>> futures =
|
||||
exec.invokeAll(tasks);
|
||||
Integer sum = futures.stream()
|
||||
.map(CachedThreadPool3::extractResult)
|
||||
.reduce(0, Integer::sum);
|
||||
System.out.println("sum = " + sum);
|
||||
exec.shutdown();
|
||||
}
|
||||
}
|
||||
/* Output:
|
||||
1 pool-1-thread-2 100
|
||||
0 pool-1-thread-1 100
|
||||
4 pool-1-thread-5 100
|
||||
5 pool-1-thread-6 100
|
||||
8 pool-1-thread-9 100
|
||||
9 pool-1-thread-10 100
|
||||
2 pool-1-thread-3 100
|
||||
3 pool-1-thread-4 100
|
||||
6 pool-1-thread-7 100
|
||||
7 pool-1-thread-8 100
|
||||
sum = 1000
|
||||
*/
|
||||
|
||||
```
|
||||
|
||||
只有在所有任务完成后,**invokeAll()**才会返回一个**Future**列表,每个任务一个**Future**。**Future**是Java 5中引入的机制,允许您提交任务而无需等待它完成。在这里,我们使用**ExecutorService.submit()**:
|
||||
|
||||
```java
|
||||
// concurrent/Futures.java
|
||||
import java.util.*;
|
||||
import java.util.concurrent.*;
|
||||
import java.util.stream.*;
|
||||
public class Futures {
|
||||
public static void main(String[] args)throws InterruptedException, ExecutionException {
|
||||
ExecutorService exec
|
||||
=Executors.newSingleThreadExecutor();
|
||||
Future<Integer> f =
|
||||
exec.submit(newCountingTask(99));
|
||||
System.out.println(f.get()); // [1]
|
||||
exec.shutdown();
|
||||
}
|
||||
}
|
||||
/* Output:
|
||||
99 pool-1-thread-1 100
|
||||
100
|
||||
*/
|
||||
```
|
||||
|
||||
- [1] 当你的任务尚未完成的**Future**上调用**get()**时,调用会阻塞(等待)直到结果可用。
|
||||
|
||||
但这意味着,在**CachedThreadPool3.java**中,**Future**似乎是多余的,因为**invokeAll()**甚至在所有任务完成之前都不会返回。但是,这里的Future并不用于延迟结果,而是用于捕获任何可能发生的异常。
|
||||
|
||||
还要注意在**CachedThreadPool3.java.get()**中抛出异常,因此**extractResult()**在Stream中执行此提取。
|
||||
|
||||
因为当你调用**get()**时,**Future**会阻塞,所以它只能解决等待任务完成的问题。最终,**Futures**被认为是一种无效的解决方案,现在不鼓励,支持Java 8的**CompletableFuture**,我们将在本章后面探讨。当然,您仍会在遗留库中遇到Futures
|
||||
|
||||
我们可以使用并行Stream以更简单,更优雅的方式解决这个问题:
|
||||
|
||||
```java
|
||||
// concurrent/CountingStream.java
|
||||
// {VisuallyInspectOutput}
|
||||
import java.util.*;
|
||||
import java.util.concurrent.*;
|
||||
import java.util.stream.*;
|
||||
public class CountingStream {
|
||||
public static void main(String[] args) {
|
||||
System.out.println(
|
||||
IntStream.range(0, 10)
|
||||
.parallel()
|
||||
.mapToObj(CountingTask::new)
|
||||
.map(ct -> ct.call())
|
||||
.reduce(0, Integer::sum));
|
||||
}
|
||||
}
|
||||
/* Output:
|
||||
1 ForkJoinPool.commonPool-worker-3 100
|
||||
8 ForkJoinPool.commonPool-worker-2 100
|
||||
0 ForkJoinPool.commonPool-worker-6 100
|
||||
2 ForkJoinPool.commonPool-worker-1 100
|
||||
4 ForkJoinPool.commonPool-worker-5 100
|
||||
9 ForkJoinPool.commonPool-worker-7 100
|
||||
6 main 100
|
||||
7 ForkJoinPool.commonPool-worker-4 100
|
||||
5 ForkJoinPool.commonPool-worker-2 100
|
||||
3 ForkJoinPool.commonPool-worker-3 100
|
||||
1000
|
||||
*/
|
||||
```
|
||||
|
||||
这不仅更容易理解,我们需要做的就是将**parallel()**插入到其他顺序操作中,然后一切都在同时运行。
|
||||
|
||||
- Lambda和方法引用作为任务
|
||||
|
||||
使用lambdas和方法引用,您不仅限于使用**Runnables**和**Callables**。因为Java 8通过匹配签名来支持lambda和方法引用(即,它支持结构一致性),所以我们可以将notRunnables或Callables的参数传递给ExecutorService:
|
||||
|
||||
使用lambdas和方法引用,您不仅限于使用**Runnables**和**Callables**。因为Java 8通过匹配签名来支持lambda和方法引用(即,它支持结构一致性),所以我们可以将不是**Runnables**或**Callables**的参数传递给**ExecutorService**:
|
||||
|
||||
```java
|
||||
// concurrent/LambdasAndMethodReferences.java
|
||||
import java.util.concurrent.*;
|
||||
class NotRunnable {
|
||||
public void go() {
|
||||
System.out.println("NotRunnable");
|
||||
}
|
||||
}
|
||||
class NotCallable {
|
||||
public Integer get() {
|
||||
System.out.println("NotCallable");
|
||||
return 1;
|
||||
}
|
||||
}
|
||||
public class LambdasAndMethodReferences {
|
||||
public static void main(String[] args)throws InterruptedException {
|
||||
ExecutorService exec =
|
||||
Executors.newCachedThreadPool();
|
||||
exec.submit(() -> System.out.println("Lambda1"));
|
||||
exec.submit(newNotRunnable()::go);
|
||||
exec.submit(() -> {
|
||||
System.out.println("Lambda2");
|
||||
return 1;
|
||||
});
|
||||
exec.submit(newNotCallable()::get);
|
||||
exec.shutdown();
|
||||
}
|
||||
}
|
||||
/* Output:
|
||||
Lambda1
|
||||
NotCallable
|
||||
NotRunnable
|
||||
Lambda2
|
||||
*/
|
||||
|
||||
```
|
||||
|
||||
这里,前两个**submit()**调用可以改为调用**execute()**。所有**submit()**调用都返回**Futures**,您可以在后两次调用的情况下提取结果。
|
||||
<!-- Terminating Long-Running Tasks -->
|
||||
## 终止耗时任务
|
||||
|
||||
|
||||
Reference in New Issue
Block a user