Java 线程池详解

网友投稿 217 2023-07-19


Java 线程池详解

系统启动一个线程的成本是比较高的,因为它涉及到与操作系统的交互,使用线程池的好处是提高性能,当系统中包含大量并发的线程时,会导致系统性能剧烈下降,甚至导致JVM崩溃,而线程池的最大线程数参数可以控制系统中并发线程数不超过次数。

一、Executors 工厂类用来产生线程池,该工厂类包含以下几个静态工厂方法来创建对应的线程池。创建的线程池是一个ExecutorService对象,使用该对象的submit方法或者是execute方法执行相应的Runnable或者是Callable任务。线程池本身在不再需要的时候调用shutdown()方法停止线程池,调用该方法后,该线程池将不再允许任务添加进来,但是会直到已添加的所有任务执行完成后才死亡。

1、newCachedThreadPool(),创建一个具有缓存功能的线程池,提交到该线程池的任务(Runnable或Callable对象)创建的线程,如果执行完成,会被缓存到CachedThreadPool中,供后面需要执行的任务使用。

import java.util.concurrent.ExecutorService;

import java.util.concurrent.Executors;

public class CacheThreadPool {

static class Task implements Runnable {

@Override

public void run() {

System.out.println(this + " " + Thread.currentThread().getName() + " AllStackTraces map size: "

+ Thread.currentThread().getAllStackTraces().size());

}

}

public static void main(String[] args) {

ExecutorService cacheThreadPool = Executors.newCachedThreadPool();

//先添加三个任务到线程池

for(int i = 0 ; i < 3; i++) {

cacheThreadPool.execute(new Task());

}

//等三个线程执行完成后,再次添加三个任务到线程池

try {

Thread.sleep(3000);

} catch (InterruptedException e) {

e.printStackTrace();

}

for(int i = 0 ; i < 3; i++) {

cacheThreadPool.execute(new Task());

}

}

}

执行结果如下:

CacheThreadPool$Task@2d312eb9 pool-1-thread-1 AllStackTraces map size: 7

CacheThreadPool$Task@59522b86 pool-1-thread-3 AllStackTraces map size: 7

CacheThreadPool$Task@73dbb89f pool-1-thread-2 AllStackTraces map size: 7

CacheThreadPool$Task@5795cedc pool-1-thread-3 AllStackTraces map size: 7

CacheThreadPool$Task@256d5600 pool-1-thread-1 AllStackTraces map size: 7

CacheThreadPool$Task@7d1c5894 pool-1-thread-2 AllStackTraces map size: 7

线程池中的线程对象进行了缓存,当有新任务执行时进行了复用。但是如果有特别多的并发时,缓存线程池还是会创建很多个线程对象。

2、newFixedThreadPool(int nThreads) 创建一个指定线程个数,线程可复用的线程池。

import java.util.concurrent.ExecutorService;

import java.util.concurrent.Executors;

public class FixedThreadPool {

static class Task implements Runnable {

@Override

public void run() {

System.out.println(this + " " + Thread.currentThread().getName() + " AllStackTraces map size: "

+ Thread.currentThread().getAllStackTraces().size());

}

}

public static void main(String[] args) {

ExecutorService fixedThreadPool = Executors.newFixedThreadPool(3);

// 先添加三个任务到线程池

for (int i = 0; i < 5; i++) {

fixedThreadPool.execute(new Task());

}

// 等三个线程执行完成后,再次添加三个任务到线程池

try {

Thread.sleep(3);

} catch (InterruptedException e) {

e.printStackTrace();

}

for (int i = 0; i < 3; i++) {

fixedThreadPool.execute(new Task());

}

}

}

执行结果:

FixedThreadPool$Task@7045c12d pool-1-thread-2 AllStackTraces map size: 7

FixedThreadPool$Task@50fa0bef pool-1-thread-2 AllStackTraces map size: 7

FixedThreadPool$Task@ccb1870 pool-1-thread-2 AllStackTraces map size: 7

FixedThreadPool$Task@7392b4e3 pool-1-thread-1 AllStackTraces map size: 7

FixedThreadPool$Task@5bdeff18 pool-1-thread-2 AllStackTraces map size: 7

FixedThreadPool$Task@7d5554e1 pool-1-thread-1 AllStackTraces map sizehttp://: 7

FixedThreadPool$Task@24468092 pool-1-thread-3 AllStackTraces map size: 7

FixedThreadPool$Task@fa7b978 pool-1-thread-2 AllStackTraces map size: 7

3、newSingleThreadExecutor(),创建一个只有单线程的线程池,相当于调用newFixedThreadPool(1)

4、newSheduledThreadPool(int corePoolSize),创建指定线程数的线程池,它可以在指定延迟后执行线程。也可以以某一周期重复执行某一线程,知道调用shutdown()关闭线程池。

示例如下:

import java.util.concurrent.Executors;

import java.util.concurrent.ScheduledExecutorService;

import java.util.concurrent.TimeUnit;

public class ScheduledThreadPool {

static class Task implements Runnable {

@Override

public void run() {

System.out.println("time " + System.currentTimeMillis() + " " + Thread.currentThread().getName() + " AllStackTraces map size: "

+ Thread.currentThread().getAllStackTraces().size());

}

}

public static void main(String[] args) {

ScheduledExecutorService scheduledExecutorService = Executors.newScheduledThreadPool(3);

scheduledExecutorService.schedule(new Task(), 3, TimeUnit.SECONDS);

scheduledExecutorService.scheduleAtFixedRate(new Task(), 3, 5, TimeUnit.SECONDS);

try {

Thread.sleep(30 * 1000);

} catch (InterruptedException e) {

e.printStackTrace();

}

scheduledExecutorService.shutdown();

}

}

运行结果如下:

time 1458921795240 pool-1-thread-1 AllStackTraces map size: 6

time 1458921795241 pool-1-thread-2 AllStackTraces map size: 6

time 1458921800240 pool-1-thread-1 AllStackTraces map size: 7

time 1458921805240 pool-1-thread-1 AllStackTraces map size: 7

time 1458921810240 pool-1-thread-1 AllStackTraces map size: 7

time 1458921815240 pool-1-thread-1 AllStackTraces map size: 7

time 1458921820240 pool-1-thread-1 AllStackTraces map size: 7

由运行时间可看出,任务是按照5秒的周期执行的。

5、newSingleThreadScheduledExecutor() 创建一个只有一个线程的线程池,同调用newScheduledThreadPool(1)。

二、ForkJoinPool和ForkJoinTask

ForkJoinPool是ExecutorService的实现类,支持将一个任务划分为多个小任务并行计算,在把多个小任务的计算结果合并成总的计算结果。它有两个构造函数

ForkJoinPool(int parallelism)创建一个包含parallelism个并行线程的ForkJoinPool。

ForkJoinPool(),以Runtime.availableProcessors()方法返回值作为parallelism参数来创建ForkJoinPool。

ForkJoinTask 代表一个可以并行,合并的任务。它是实现了Future接口的抽象类,它有两个抽象子类,代表无返回值任务的RecuriveAction和有返回值的RecursiveTask。可根据具体需求继承这两个抽象类实现自己的对象,然后调用ForkJoinPool的submit 方法执行。

RecuriveAction 示例如下,实现并行输出0-300的数字。

import java.util.concurrent.ForkJoinPool;

import java.util.concurrent.RecursiveAction;

import java.util.concurrent.TimeUnit;

public clqLiOfuKJcass ActionForkJoinTask {

static class PrintTask extends RecursiveAction {

private static final int THRESHOLD = 50;

private int start;

private int end;

public PrintTask(int start, int end) {

this.start = start;

this.end = end;

}

@Override

protected void compute() {

if (end - start < THRESHOLD) {

for(int i = start; i < end; i++) {

System.out.println(Thread.currentThread().getName() + " " + i);

}

} else {

int middle = (start + end) / 2;

PrintTask left = new PrintTask(start, middle);

PrintTask right = new PrintTask(middle, end);

left.fork();

right.fork();

}

}

}

public static void main(String[] args) {

ForkJoinPool pool = new ForkJoinPool();

pool.submit(new PrintTask(0, 300));

try {

pool.awaitTermination(2, TimeUnit.SECONDS);

} catch (InterruptedException e) {

e.printStackTrace();

}

pool.shutdown();

}

}

在拆分小任务后,调用任务的fork()方法,加入到ForkJoinPool中并行执行。

RecursiveTask示例,实现并行计算100个整数求和。拆分为每20个数求和后获取结果,在最后合并为最后的结果。

import java.util.Random;

import java.util.concurrent.ExecutionException;

import java.util.concurrent.ForkJoinPool;

import java.util.concurrent.Future;

import java.util.concurrenhttp://t.RecursiveTask;

public class TaskForkJoinTask {

static class CalTask extends RecursiveTask {

private static final int THRESHOLD = 20;

private int arr[];

private int start;

private int end;

public CalTask(int[] arr, int start, int end) {

this.arr = arr;

this.start = start;

this.end = end;

}

@Override

protected Integer compute() {

int sum = 0;

if (end - start < THRESHOLD) {

for (int i = start; i < end; i++) {

sum += arr[i];

}

System.out.println(Thread.currentThread().getName() + " sum:" + sum);

return sum;

} else {

int middle = (start + end) / 2;

CalTask left = new CalTask(arr, start, middle);

CalTask right = new CalTask(arr, middle, end);

left.fork();

right.fork();

return left.join() + right.join();

}

}

}

public static void main(String[] args) {

int arr[] = new int[100];

Random random = new Random();

int total = 0;

for (int i = 0; i < arr.length; i++) {

int tmp = random.nextInt(20);

total += (arr[i] = tmp);

}

System.out.println("total " + total);

ForkJoinPool pool = new ForkJoinPool(4);

Future future = pool.submit(new CalTask(arr, 0, arr.length));

try {

System.out.println("cal result: " + future.get());

} catch (InterruptedException e) {

e.printStackTrace();

} catch (ExecutionException e) {

e.printStackTrace();

}

pool.shutdown();

}

}

执行结果如下:

total 912

ForkJoinPool-1-worker-2 sum:82

ForkJoinPool-1-worker-2 sum:123

ForkJoinPool-1-worker-2 sum:144

ForkJoinPool-1-worker-3 sum:119

ForkJoinPool-1-worker-2 sum:106

ForkJoinPool-1-worker-2 sum:128

ForkJoinPool-1-worker-2 sum:121

ForkJoinPool-1-worker-3 sum:89

cal result: 912

子任务执行完后,调用任务的join()方法获取子任务执行结果,再相加获得最后的结果。


版权声明:本文内容由网络用户投稿,版权归原作者所有,本站不拥有其著作权,亦不承担相应法律责任。如果您发现本站中有涉嫌抄袭或描述失实的内容,请联系我们jiasou666@gmail.com 处理,核实后本网站将在24小时内删除侵权内容。

上一篇:Java 中ThreadLocal类详解
下一篇:Hibernate延迟加载原理与实现方法
相关文章

 发表评论

暂时没有评论,来抢沙发吧~