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 size: 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<T>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 class 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.concurrent.RecursiveTask; public class TaskForkJoinTask {
static class CalTask extends RecursiveTask<Integer> {
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<Integer> 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。