用JAVA實現(xiàn)一套背壓機制

更新時間：2023年06月30日 08:47:27 作者：hwp0710

背壓依我的理解來說，是指訂閱者能和發(fā)布者交互，可以調節(jié)發(fā)布者發(fā)布數(shù)據(jù)的速率，解決把訂閱者壓垮的問題，這篇文章主要介紹了用JAVA自己實現(xiàn)一套背壓機制,需要的朋友可以參考下

Reactive Streams：一種支持背壓的異步數(shù)據(jù)流處理標準，主流實現(xiàn)有RxJava和Reactor，Spring WebFlux默認集成的是Reactor。

Reactive Streams主要解決背壓(back-pressure)問題。當傳入的任務速率大于系統(tǒng)處理能力時，數(shù)據(jù)處理將會對未處理數(shù)據(jù)產(chǎn)生一個緩沖區(qū)。

背壓依我的理解來說，是指訂閱者能和發(fā)布者交互(通過代碼里面的調用request和cancel方法交互)，可以調節(jié)發(fā)布者發(fā)布數(shù)據(jù)的速率，解決把訂閱者壓垮的問題。關鍵在于上面例子里面的訂閱關系Subscription這個接口，他有request和cancel 2個方法，用于通知發(fā)布者需要數(shù)據(jù)和通知發(fā)布者不再接受數(shù)據(jù)。

我們重點理解背壓在jdk9里面是如何實現(xiàn)的。關鍵在于發(fā)布者Publisher的實現(xiàn)類SubmissionPublisher的submit方法是block方法。訂閱者會有一個緩沖池，默認為Flow.defaultBufferSize() = 256。當訂閱者的緩沖池滿了之后，發(fā)布者調用submit方法發(fā)布數(shù)據(jù)就會被阻塞，發(fā)布者就會停(慢)下來；訂閱者消費了數(shù)據(jù)之后(調用Subscription.request方法)，緩沖池有位置了，submit方法就會繼續(xù)執(zhí)行下去，就是通過這樣的機制，實現(xiàn)了調節(jié)發(fā)布者發(fā)布數(shù)據(jù)的速率，消費得快，生成就快，消費得慢，發(fā)布者就會被阻塞，當然就會慢下來了。

單線程版本：

一個生產(chǎn)者，一個消費者

import lombok.SneakyThrows;
import java.util.ArrayList;
import java.util.List;
import java.util.Random;
public class BackpressureExample {
    public static void main(String[] args) throws InterruptedException {
        BackpressureSubscriber subscriber = new BackpressureSubscriber();
        BackpressurePublisher publisher = new BackpressurePublisher(subscriber);
        publisher.start();
        subscriber.start();
        // 為了演示效果，這里讓主線程休眠一段時間
        Thread.sleep(50000);
        publisher.stop();
        subscriber.stop();
    }
    @SneakyThrows
    public static void processDataLogic(List<Integer> batch) {
        //模擬任務執(zhí)行
        int r = new Random().nextInt(3000);
        Thread.sleep(r);
        System.out.println(Thread.currentThread().getName() + ",Received batch: " + batch + ",sleep ms = " + r);
    }
    static class BackpressurePublisher {
        private final BackpressureSubscriber subscriber;
        private volatile boolean running;
        public BackpressurePublisher(BackpressureSubscriber subscriber) {
            this.subscriber = subscriber;
            this.running = true;
        }
        public void start() {
            Thread thread = new Thread(() -> {
                int item = 1;
                while (running) {
                    List<Integer> batch = new ArrayList<>();
                    for (int i = 0; i < 5; i++) {
                        System.out.println(Thread.currentThread().getName() + "-----produce data = " + item);
                        batch.add(item++);
                    }
                    while (!subscriber.accept(batch)) {
                        if (!running) {
                            break;
                        }
                    }
                }
            });
            thread.start();
        }
        public void stop() {
            running = false;
        }
    }
    static class BackpressureSubscriber {
        private volatile boolean running;
        public BackpressureSubscriber() {
            this.running = true;
        }
        public boolean accept(List<Integer> batch) {
            if (running) {
                processDataLogic(batch);
                return true;
            } else {
                return false;
            }
        }
        public void start() {
            // Subscriber 在 JDK 8 中沒有異步處理的能力，因此不需要單獨開啟線程
        }
        public void stop() {
            running = false;
        }
    }
}

多線程版本

一個生產(chǎn)者，多個消費者

import lombok.SneakyThrows;
import java.util.ArrayList;
import java.util.List;
import java.util.Random;
import java.util.concurrent.ExecutorService;
import java.util.concurrent.Executors;
import java.util.concurrent.Future;
public class BackpressureExample {
    public static void main(String[] args) throws InterruptedException {
        BackpressureSubscriber subscriber = new BackpressureSubscriber();
        BackpressurePublisher publisher = new BackpressurePublisher(subscriber);
        publisher.start();
        subscriber.start();
        // 為了演示效果，這里讓主線程休眠一段時間
        Thread.sleep(50000);
        publisher.stop();
        subscriber.stop();
    }
    @SneakyThrows
    public static void processDataLogic(List<Integer> batch) {
        //模擬任務執(zhí)行
        int r = new Random().nextInt(3000);
        Thread.sleep(r);
        System.out.println(Thread.currentThread().getName() + ",Received batch: " + batch + ",sleep ms = " + r);
    }
    static class BackpressurePublisher {
        private final BackpressureSubscriber subscriber;
        private volatile boolean running;
        public BackpressurePublisher(BackpressureSubscriber subscriber) {
            this.subscriber = subscriber;
            this.running = true;
        }
        public void start() {
            Thread thread = new Thread(() -> {
                int item = 1;
                while (running) {
                    List<Integer> batch = new ArrayList<>();
                    for (int i = 0; i < 5; i++) {
                        System.out.println(Thread.currentThread().getName() + "-----produce data = " + item);
                        batch.add(item++);
                    }
                    while (!subscriber.accept(batch)) {
                        if (!running) {
                            break;
                        }
                    }
                }
            });
            thread.start();
        }
        public void stop() {
            running = false;
        }
    }
    static class BackpressureSubscriber {
        private volatile boolean running;
        private final ExecutorService executor;
        private final int workerSize = 2;
        private final List<Future> futures;
        public BackpressureSubscriber() {
            this.running = true;
            this.executor = Executors.newFixedThreadPool(workerSize);
            futures = new ArrayList<>(workerSize);
        }
        public boolean accept(List<Integer> batch) {
            if (running) {
                Future f = executor.submit(() -> processDataLogic(batch));
                futures.add(f);
                waitForTaskDone(futures);
                return true;
            } else {
                return false;
            }
        }
        public void waitForTaskDone(List<Future> futures) {
            while (futures.size() >= workerSize) {
                for (Future future : futures) {
                    if (future.isDone()) {
                        // 只要有一個worker是空閑就重新獲取任務
                        futures.remove(future);
                        return;
                    }
                }
            }
        }
        public void start() {
            // Subscriber 在 JDK 8 中沒有異步處理的能力，因此不需要單獨開啟線程
        }
        public void stop() {
            running = false;
            executor.shutdown();
        }
    }
}

到此這篇關于用JAVA自己實現(xiàn)一套背壓機制的文章就介紹到這了,更多相關java背壓機制內(nèi)容請搜索腳本之家以前的文章或繼續(xù)瀏覽下面的相關文章希望大家以后多多支持腳本之家！