在不断变化的过程中，只要有oom，整个过程就会陷入困境，直到我们手动杀死它。我创建了一个例子来模拟它。请参考以下代码：

@Test
public void test_Flux() {
    Flux.range(1, 30)
            .parallel(6)
            .runOn(Schedulers.fromExecutor(Executors.newFixedThreadPool(6)))
            .doOnNext(this::writeValues)
            .sequential()
            .blockLast();
}

private void writeValues(int val) {
    if (val == 10) {
        throw new OutOfMemoryError("failed here");
        // throw new RuntimeException("failed here");
    }
    System.out.println("the thread name is :" + Thread.currentThread().getName() + " val is :" + val);
    sleep(3);
}

private void sleep(int i) {
    try {
        Thread.sleep(i * 1000L);
    } catch (InterruptedException e) {
        Thread.currentThread().interrupt();
        throw new RuntimeException(e);
    }
}

我们能在这里撞车而不是被卡住吗？