目录
- 一、说明
- 二、理解
- 三、实现
- 1.AbortPolicy
- 2.DiscardPolicy
- 3.DiscardOldestPolicy
- 4.CallerRunsPolicy
- 5.自定义拒绝执行策略
一、说明
RejectedExecutionHandler
- 当线程池已经被关闭，或者任务数超过maximumPoolSize+workQueue时执行拒绝策略
-
ThreadPoolExecutor.AbortPolicy 默认拒绝策略，丢弃任务并抛出RejectedExecutionException异常 -
ThreadPoolExecutor.DiscardPolicy 直接丢弃任务，但不抛出异常 -
ThreadPoolExecutor.DiscardOldestPolicy 丢弃任务队列最先加入的任务，再执行execute方法把新任务加入队列执行 -
ThreadPoolExecutor.CallerRunsPolicy：由创建了线程池的线程来执行被拒绝的任务
二、理解
AbortPolicy
默认拒绝策略，丢弃任务并抛出RejectedExecutionException异常
private static final RejectedExecutionHandler defaultHandler =
new AbortPolicy();
public static class AbortPolicy implements RejectedExecutionHandler {r
public AbortPolicy() { }
public void rejectedExecution(Runnable r, ThreadPoolExecutor e) {
throw new RejectedExecutionException("Task " + r.toString() +
" rejected from " +
e.toString());
}
}DiscardPolicy
直接丢弃任务，但不抛出异常
public static class DiscardPolicy implements RejectedExecutionHandler {
public DiscardPolicy() { }
public void rejectedExecution(Runnable r, ThreadPoolExecutor e) {
}
}DiscardOldestPolicy
丢弃队列中等待最久的任务，再把新任务添加进去执行，从任务队列弹出最先加入的任务，空出一个位置，然后再次执行execute方法把任务加入队列
public static class DiscardOldestPolicy implements RejectedExecutionHandler {
public DiscardOldestPolicy() { }
public void rejectedExecution(Runnable r, ThreadPoolExecutor e) {
if (!e.isShutdown()) {
e.getQueue().poll();
e.execute(r);
}
}
}CallerRunsPolicy
会调用当前线程池的所在的线程去执行被拒绝的任务
public static class CallerRunsPolicy implements RejectedExecutionHandler {
public CallerRunsPolicy() { }
public void rejectedExecution(Runnable r, ThreadPoolExecutor e) {
if (!e.isShutdown()) {
r.run();
}
}
}三、实现
1.AbortPolicy
创建 ThreadPoolExecutorTest类，默认使用ThreadPoolExecutor.AbortPolicy拒绝策略，队列是ArrayBlockingQueue，设置核心线程数最大值为1，线程池线程数最大值为2，最大等待时间为5秒，等待队列值为2
public class RejectedExecutionHandlerTest {
public static void main(String[] args) throws InterruptedException {
// 1.创建自定义线程池
ThreadPoolExecutor threadPoolExecutor = new ThreadPoolExecutor(2, 4, 5,
TimeUnit.SECONDS,
new ArrayBlockingQueue<>(2),
Executors.defaultThreadFactory(),
new ThreadPoolExecutor.AbortPolicy());
// 2.创建线程任务
for (int i = 1; i <= 6; i++) {
// 3.执行任务
System.out.println("执行第"+i+"个任务");
threadPoolExecutor.execute(new runnable("任务"+i));
System.out.println("当前核心线程数" + threadPoolExecutor.getCorePoolSize());
System.out.println("当前线程池线程数" + threadPoolExecutor.getPoolSize());
// 4.迭代器获取等待队列
Iterator iterator = threadPoolExecutor.getQueue().iterator();
System.out.print("当前等待队列 ");
while (iterator.hasNext()){
runnable thread = (runnable) iterator.next();
System.out.print(thread.name + "\t");
}
System.out.print("\n");
System.out.println("--------");
}
Thread.sleep(10000);
System.out.println("----休眠10秒后----");
System.out.println("当前核心线程数" + threadPoolExecutor.getCorePoolSize());
System.out.println("当前线程池线程数" + threadPoolExecutor.getPoolSize());
System.out.println("当前队列任务数" + threadPoolExecutor.getQueue().size());
// 5.关闭线程池
threadPoolExecutor.shutdown();
}
// 实现Runnable
static class runnable implements Runnable{
// 设置任务名
String name;
public runnable(String setName) {
this.name = setName;
}
@Override
public void run() {
try {
System.out.println("线程:"+Thread.currentThread().getName() +" 执行: "+name);
} catch (Exception e) {
e.printStackTrace();
}
}
}
}当线程数达到corePoolSize后，若有新任务加入，则直接进入任务队列等待，超出队列的任务会创建新的线程来执行
一共有1个核心，当线程数超过corePoolSize+workQueue时，将创建非核心线程，核心线程默认情况下不会被回收，不受时间限制，而超时的非核心线程将被回收
但如果再执行1个任务，线程数超过maximumPoolSize+workQueue，再提交任务将被丢弃并抛出RejectedExecutionException异常
2.DiscardPolicy
创建5个任务，让被线程池拒绝的任务直接丢弃，不会抛异常也不会执行
ThreadPoolExecutor threadPoolExecutor = new ThreadPoolExecutor(1, 2, 5,
TimeUnit.SECONDS,
new ArrayBlockingQueue<>(2),
Executors.defaultThreadFactory(),
new ThreadPoolExecutor.DiscardPolicy());
任务5不会执行，恶意不会抛出异常，超时的非核心线程将被回收
3.DiscardOldestPolicy
丢弃任务队列最先加入的任务，再执行execute方法把新任务加入队列执行
ThreadPoolExecutor threadPoolExecutor = new ThreadPoolExecutor(1, 2, 5,
TimeUnit.SECONDS,
new ArrayBlockingQueue<>(2),
Executors.defaultThreadFactory(),
new ThreadPoolExecutor.DiscardOldestPolicy());
添加任务5时，线程数已经超过maximumPoolSize+workQueue，抛弃最先加入队列的任务2并且不执行，再将任务5加进队列中执行
4.CallerRunsPolicy
会调用当前线程池的所在的线程去执行被拒绝的任务
// 1.创建自定义线程池
ThreadPoolExecutor threadPoolExecutor = new ThreadPoolExecutor(1, 2, 5,
TimeUnit.SECONDS,
new ArrayBlockingQueue<>(2),
Executors.defaultThreadFactory(),
new ThreadPoolExecutor.CallerRunsPolicy());
主线程执行任务1，空闲线程执行任务4，此时队列中有任务2和任务3
添加任务5时，线程数已经超过maximumPoolSize+workQueue，任务5直接调用当前线程池的所在的线程main去执行，这时主线程被阻塞了
当任务5执行完成时，最先的两个任务已经完成了，主线程去执行任务2和任务3，添加任务6也可以直接执行
超时的非核心线程将被回收
5.自定义拒绝执行策略
当线程数已经超过maximumPoolSize+workQueue时，调用新线程去执行任务
static class MyRejectedExecutionHandler implements RejectedExecutionHandler {
@Override
public void rejectedExecution(Runnable r, ThreadPoolExecutor executor) {
new Thread(r, "新线程"+(new Random().nextInt(4) + 1)).start();
}
}ThreadPoolExecutor threadPoolExecutor = new ThreadPoolExecutor(1, 2, 5,
TimeUnit.SECONDS,
new ArrayBlockingQueue<>(2),
Executors.defaultThreadFactory(),
new MyRejectedExecutionHandler());
