RCountDownLatch 分布式计数器锁的使用示例

RCountDownLatch 是 Redisson 提供的一种分布式计数器锁，类似于 Java 的 CountDownLatch。

它允许一个或多个线程等待其他操作完成后再执行，适用于分布式环境中需要协调多任务的场景。

以下示例设计来自ChatGPT。

1.示例场景

假设有 5 个任务，主线程需要等这 5 个任务全部完成后再继续执行。

public static void main(String[] args) {
    // Create a Config instance and configure Redis connection
    Config config = new Config();
    config.useSingleServer().setAddress("redis://" + REDIS_HOST + ":" + REDIS_PORT).setPassword(redisPassword);
    // Create a RedissonClient instance
    RedissonClient redissonClient = Redisson.create(config);
    // Create a RCountDownLatch instance
    RCountDownLatch latch = redissonClient.getCountDownLatch("myCountDownLatch");

    // Set the initial count of the latch
    latch.trySetCount(5);

    // Start the worker thread and wait for the count to reset to zero
    new Thread(new Worker(redissonClient)).start();
    new Thread(new Worker(redissonClient)).start();
    new Thread(new Worker(redissonClient)).start();
    new Thread(new Worker(redissonClient)).start();
    new Thread(new Worker(redissonClient)).start();

    try {
        // The main thread is waiting for other threads to complete
        latch.await();
        System.out.println("All tasks have been completed, and the main thread continues to execute");
    } catch (InterruptedException e) {
        e.printStackTrace();
    } finally {
        redissonClient.shutdown();
    }
    System.out.println("end this demo!");
}

// Worker class that implements Runnable interface
static class Worker implements Runnable {
    private final RedissonClient redissonClient;

    public Worker(RedissonClient redissonClient) {
        this.redissonClient = redissonClient;
    }

    @Override
    public void run() {
        RCountDownLatch latch = redissonClient.getCountDownLatch("myCountDownLatch");

        try {
            System.out.println("Task " + Thread.currentThread().getName() + " start");
            Thread.sleep((int) (Math.random() * 3000));  // Simulate task execution
            System.out.println("Task " + Thread.currentThread().getName() + " completed");
        } catch (InterruptedException e) {
            e.printStackTrace();
        } finally {
            latch.countDown();  // After each task is completed, the count is reduced by one
        }
    }
}

> Task :CountDownLatchExample.main()
Task Thread-1 start
Task Thread-2 start
Task Thread-3 start
Task Thread-5 start
Task Thread-4 start
Task Thread-4 completed
Task Thread-3 completed
Task Thread-1 completed
Task Thread-5 completed
Task Thread-2 completed
All tasks have been completed, and the main thread continues to execute
end this demo!

2. 使用 RCountDownLatch 在分布式系统中进行任务同步

在分布式系统中，多个 JVM 上的线程也可以通过共享的 RCountDownLatch 对象来同步任务。

只要使用相同的 Redis 服务器和相同的锁名称，就可以在多个应用实例中同步使用 RCountDownLatch。