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。
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
|
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:
1
2
3
4
5
6
|
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。
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
|
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:
1
2
3
4
5
6
7
8
|
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:
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
|
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:
1
2
3
4
5
6
7
|
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。
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
|
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。
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
|
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:
1
2
3
4
5
6
7
8
9
10
|
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。