Guava RateLimiter限流

版权声明：本文为博主原创文章，未经博主允许不得转载。 https://blog.csdn.net/u013278314/article/details/82263071

RateLimiter是guava提供的基于令牌桶算法的实现类，可以非常简单的完成限流特技，并且根据系统的实际情况来调整生成token的速率。

通常可应用于抢购限流防止冲垮系统；限制某接口、服务单位时间内的访问量，譬如一些第三方服务会对用户访问量进行限制；限制网速，单位时间内只允许上传下载多少字节等。

下面是限制每秒钟最大能有10个获取token:

import java.util.Random;
import java.util.concurrent.CountDownLatch;
import java.util.concurrent.ExecutorService;
import java.util.concurrent.Executors;
import java.util.concurrent.TimeUnit;

import com.google.common.util.concurrent.RateLimiter;

public class Test {

	RateLimiter rateLimiter = RateLimiter.create(10);

	public static void main(String[] args) {
		Test t = new Test();

		ExecutorService s = Executors.newFixedThreadPool(20);
		CountDownLatch countDownLatch = new CountDownLatch(1);

		Random random = new Random(10);

		for (int i = 0; i < 20; i++) {
			s.submit(new Runnable() {

				@Override
				public void run() {

					try {
						countDownLatch.await();
						TimeUnit.MILLISECONDS.sleep(random.nextInt(1000));
						t.seckill();
					} catch (InterruptedException e1) {
						e1.printStackTrace();
					}

				}
			});
		}
		countDownLatch.countDown();

		s.shutdown();
	}

	public void seckill() {

		if (rateLimiter.tryAcquire()) {
			System.out.println(Thread.currentThread().getName() + "： 我抢到了");
		} else {
			System.out.println(Thread.currentThread().getName() + "： 我没抢到");
		}

	}

}

运行结果：

下面是限制每秒钟只能获取一个token,20个线程需要20秒才能获取完

RateLimiter rateLimiter = RateLimiter.create(10);

	public static void main(String[] args) {
		Test t = new Test();

		ExecutorService s = Executors.newFixedThreadPool(20);
		CountDownLatch countDownLatch = new CountDownLatch(1);

		for (int i = 0; i < 20; i++) {
			s.submit(new Runnable() {

				@Override
				public void run() {

					try {
						countDownLatch.await();
						t.seckill();
					} catch (InterruptedException e1) {
						e1.printStackTrace();
					}

				}
			});
		}
		countDownLatch.countDown();

		s.shutdown();
	}

	public void seckill() {
		System.out.println("我抢到了，耗时： " + rateLimiter.acquire() );

	}

运行结果：