1. 程式人生 > >【原始碼】RingBuffer(二)——消費者

【原始碼】RingBuffer(二)——消費者

消費者如何讀取資料?

前一篇是生產者的處理,這一篇講消費者的處理

我們都知道,消費者無非就是不停地從佇列中讀取資料,處理資料。但是與BlockedQueue不同的是,RingBuffer的消費者不會對佇列進行上鎖,那它是怎樣實現的呢?

概括地說,就是通過CAS原子性地得到一個可消費的序號,然後再根據序號取出資料進行處理。

在看程式碼之前,我們先把能想到的東西先羅列一下:

1.需要一個尾指標來追蹤消費狀態

2.如何防止一個數據被多個消費者重複消費?

3.消費速度不能超過生產者,如何限制?

4.當沒有可處理資料的時候消費者該做什麼,自旋還是掛起等待生產者喚醒?

5.如果4選擇掛起,那麼如果RingBuffer關閉,如何喚醒消費者以終結執行緒任務?

6.RingBuffer構造的時候需要傳入執行緒工廠,RingBuffer是如何使用執行緒的,多個任務使用一個執行緒排程?

7.消費者何時啟動?

好,問題有了,現在我們來看程式碼,下面是EventProcessor的一個實現,WorkProcessor的程式碼。

public final class WorkProcessor<T>
    implements EventProcessor
{
    private final AtomicBoolean running = new AtomicBoolean(false); //當前處理器狀態
    private final Sequence sequence = new Sequence(Sequencer.INITIAL_CURSOR_VALUE); //當前已消費過的最新序號
    private final RingBuffer<T> ringBuffer; //保留此引用,方便取資料
    private final SequenceBarrier sequenceBarrier; //用於等待下一個最大可用序號,可與多個Processor共用
    private final WorkHandler<? super T> workHandler; //實際上的處理器
    private final ExceptionHandler<? super T> exceptionHandler;
    private final Sequence workSequence; //多個Processor共用的workSequence,可以得到下一個待處理的序號

    private final EventReleaser eventReleaser = new EventReleaser()
    {
        @Override
        public void release()
        {
            sequence.set(Long.MAX_VALUE);
        }
    };

    private final TimeoutHandler timeoutHandler;

    /**
     * Construct a {@link WorkProcessor}.
     *
     * @param ringBuffer       to which events are published.
     * @param sequenceBarrier  on which it is waiting.
     * @param workHandler      is the delegate to which events are dispatched.
     * @param exceptionHandler to be called back when an error occurs
     * @param workSequence     from which to claim the next event to be worked on.  It should always be initialised
     *                         as {@link Sequencer#INITIAL_CURSOR_VALUE}
     */
    public WorkProcessor(
        final RingBuffer<T> ringBuffer,
        final SequenceBarrier sequenceBarrier,
        final WorkHandler<? super T> workHandler,
        final ExceptionHandler<? super T> exceptionHandler,
        final Sequence workSequence)
    {
        this.ringBuffer = ringBuffer;
        this.sequenceBarrier = sequenceBarrier;
        this.workHandler = workHandler;
        this.exceptionHandler = exceptionHandler;
        this.workSequence = workSequence;

        if (this.workHandler instanceof EventReleaseAware)
        {
            ((EventReleaseAware) this.workHandler).setEventReleaser(eventReleaser);
        }

        timeoutHandler = (workHandler instanceof TimeoutHandler) ? (TimeoutHandler) workHandler : null;
    }

    @Override
    public Sequence getSequence()
    {
        return sequence;
    }

    @Override
    public void halt()
    {
        running.set(false);
        sequenceBarrier.alert(); //喚醒卡在WaitStrategy的Processor執行緒,好讓它知道“結束”狀態
    }

    /**
     * remove workProcessor dynamic without message lost
     */
    public void haltLater()
    {
        running.set(false);  //所謂later,就是等待下一次檢查的時候才推出,如果處於卡在WaitStrategy,則等待它返回後才檢查
    }

    @Override
    public boolean isRunning()
    {
        return running.get();
    }

    /**
     * It is ok to have another thread re-run this method after a halt().
     *
     * @throws IllegalStateException if this processor is already running
     */
    @Override
    public void run()
    {
        if (!running.compareAndSet(false, true))  //防止run方法重複呼叫造成的問題
        {
            throw new IllegalStateException("Thread is already running");
        }
        sequenceBarrier.clearAlert();

        notifyStart();

        boolean processedSequence = true;
        long cachedAvailableSequence = Long.MIN_VALUE;
        long nextSequence = sequence.get();
        T event = null;
        while (true)  //死迴圈
        {
            try
            {
                // if previous sequence was processed - fetch the next sequence and set
                // that we have successfully processed the previous sequence
                // typically, this will be true
                // this prevents the sequence getting too far forward if an exception
                // is thrown from the WorkHandler

                if (processedSequence)
                {
                    if (!running.get()) //如果檢查到已關閉,則喚醒在同一個Barrier上的其他processor執行緒
                    {
                        sequenceBarrier.alert();  //喚醒其他執行緒
                        sequenceBarrier.checkAlert(); //丟擲異常,終結此執行緒
                    }
                    processedSequence = false;
                    do
                    {
                        //workSequence可能和多個Processor共用
                        nextSequence = workSequence.get() + 1L;
                        //這個sequence才是當前處理器處理過的序號,生產者判斷尾指標的時候就是按照這個來的,這個就是gatingSequence
                        //拿到下一個新序號的時候,說明workSequence前一個數據已經處理過了
                        sequence.set(nextSequence - 1L);
                    }
                    //由於workSequence可能由多個Processor共用,故存在競爭情況,需要使用CAS
                    while (!workSequence.compareAndSet(nextSequence - 1L, nextSequence));
                }

                //如果沒有超過生產者的最大遊標,則表明資料可取
                if (cachedAvailableSequence >= nextSequence)
                {
                    //取出序號對應位置的資料
                    event = ringBuffer.get(nextSequence);
                    //交給handler處理
                    workHandler.onEvent(event);
                    processedSequence = true;
                }
                else
                {
                    //阻塞等待下一個可用的序號
                    //如果就是nextSequence,就返回nextSequence
                    //如果可用的大於nextSequence,則返回最新可用的sequence
                    cachedAvailableSequence = sequenceBarrier.waitFor(nextSequence);
                }
            }
            catch (final TimeoutException e)
            {
                notifyTimeout(sequence.get());
            }
            catch (final AlertException ex) //checkAlert()丟擲的
            {
                if (!running.get()) //如果已經結束,則終結迴圈,執行緒任務結束
                {
                    break;
                }
            }
            catch (final Throwable ex) //其他異常,則交給異常處理器處理
            {
                // handle, mark as processed, unless the exception handler threw an exception
                exceptionHandler.handleEventException(ex, nextSequence, event);
                processedSequence = true;
            }
        }

        notifyShutdown();

        running.set(false);
    }

    private void notifyTimeout(final long availableSequence)
    {
        try
        {
            if (timeoutHandler != null)
            {
                timeoutHandler.onTimeout(availableSequence);
            }
        }
        catch (Throwable e)
        {
            exceptionHandler.handleEventException(e, availableSequence, null);
        }
    }

    private void notifyStart()
    {
        if (workHandler instanceof LifecycleAware)
        {
            try
            {
                ((LifecycleAware) workHandler).onStart();
            }
            catch (final Throwable ex)
            {
                exceptionHandler.handleOnStartException(ex);
            }
        }
    }

    private void notifyShutdown()
    {
        if (workHandler instanceof LifecycleAware)
        {
            try
            {
                ((LifecycleAware) workHandler).onShutdown();
            }
            catch (final Throwable ex)
            {
                exceptionHandler.handleOnShutdownException(ex);
            }
        }
    }
}

 針對問題一:需要一個尾指標來追蹤消費狀態

你們注意到程式碼中有兩個Sequence,workSequence和sequence。為啥需要兩個呢?

workSequence消費者使用的最新序號(該序號的資料未被處理過,只是被消費者標記成可消費);而sequence序號的資料則是被消費過的,這個序號正是前一篇中的gatingSequence。

針對問題二:如何防止一個數據被多個消費者重複消費?

問題二的解決方案就是WorkPool,即讓多個WorkProcessor共用一個workSequence,這樣它們就會競爭序號,一個序號只能被消費一次。

public final class WorkerPool<T>
{
    private final AtomicBoolean started = new AtomicBoolean(false);
    private final Sequence workSequence = new Sequence(Sequencer.INITIAL_CURSOR_VALUE); //從-1開始
    private final RingBuffer<T> ringBuffer; //RingBuffer引用,用於構造Processor,取資料
    // WorkProcessors are created to wrap each of the provided WorkHandlers
    private final WorkProcessor<?>[] workProcessors;
    //...
    public WorkerPool(
        final RingBuffer<T> ringBuffer,
        final SequenceBarrier sequenceBarrier,
        final ExceptionHandler<? super T> exceptionHandler,
        final WorkHandler<? super T>... workHandlers)
    {
        this.ringBuffer = ringBuffer;
        final int numWorkers = workHandlers.length;
        workProcessors = new WorkProcessor[numWorkers];

        //每個handler構造一個Processor
        for (int i = 0; i < numWorkers; i++)
        {
            workProcessors[i] = new WorkProcessor<>(
                ringBuffer,
                sequenceBarrier, //共用同一個sequenceBarrier
                workHandlers[i],
                exceptionHandler,
                workSequence); //共用同一個workSequence
        }
    }

    //...
}

public class Disruptor<T>
{
    private final RingBuffer<T> ringBuffer;
    private final Executor executor;
    private final ConsumerRepository<T> consumerRepository = new ConsumerRepository<>();
    private final AtomicBoolean started = new AtomicBoolean(false);
    private ExceptionHandler<? super T> exceptionHandler = new ExceptionHandlerWrapper<>();

    //...

    public final EventHandlerGroup<T> handleEventsWithWorkerPool(final WorkHandler<T>... workHandlers)
    {
        return createWorkerPool(new Sequence[0], workHandlers);
    }

    EventHandlerGroup<T> createWorkerPool(
        final Sequence[] barrierSequences, final WorkHandler<? super T>[] workHandlers)
    {
        final SequenceBarrier sequenceBarrier = ringBuffer.newBarrier(barrierSequences);
        final WorkerPool<T> workerPool = new WorkerPool<>(ringBuffer, sequenceBarrier, exceptionHandler, workHandlers);


        consumerRepository.add(workerPool, sequenceBarrier); //將workPool存入Repository,啟動的時候會從Repository中取出,交給Executor處理

        final Sequence[] workerSequences = workerPool.getWorkerSequences();

        updateGatingSequencesForNextInChain(barrierSequences, workerSequences);

        return new EventHandlerGroup<>(this, consumerRepository, workerSequences);
    }

    //....

}

針對問題三、四:消費速度不能超過生產者,如何限制?當沒有可處理資料的時候消費者該做什麼,自旋還是掛起等待生產者喚醒?

使用SequenceBarrier,從WorkProcessor的程式碼中我們可以知道,消費者會快取上次獲取的最大可消費序號,然後在這序號範圍內都可以直接競爭。每次獲取最小可用序號的時候,則會觸發waitStrategy等待策略進行等待。

其中等待策略有很多中,常見的就是BlockingWaitStategy,該等待策略會掛起執行執行緒。當生產者publishEvent的時候,則會呼叫WaitStrategy#signalAllWhenBlocking()方法喚醒所有等待執行緒。

 final class ProcessingSequenceBarrier implements SequenceBarrier
{
    private final WaitStrategy waitStrategy;
    private final Sequence dependentSequence;
    private volatile boolean alerted = false;
    private final Sequence cursorSequence;
    private final Sequencer sequencer;

    //...
    public long waitFor(final long sequence)
        throws AlertException, InterruptedException, TimeoutException
    {
        checkAlert();

        long availableSequence = waitStrategy.waitFor(sequence, cursorSequence, dependentSequence, this);

        if (availableSequence < sequence)
        {
            return availableSequence;
        }

        return sequencer.getHighestPublishedSequence(sequence, availableSequence);
    }
    //...
}

 針對問題六、七:RingBuffer構造的時候需要傳入執行緒工廠,RingBuffer是如何使用執行緒的,多個任務使用一個執行緒排程?消費者何時啟動?

消費者隨Disruptor啟動,Disruptor啟動時會從ConsumerRepository中取出Consumer,提交給Executor執行。

public RingBuffer<T> start()
    {
        checkOnlyStartedOnce();
        for (final ConsumerInfo consumerInfo : consumerRepository)
        {
            consumerInfo.start(executor);
        }

        return ringBuffer;
    }

其中,在新版的Disruptor中,不建議使用外部傳入的Executor,而是隻傳ThreadFactory,然後由內部構造一個Executor,就是BasicExecutor。它的實現就是每次提交的任務都建立一個新的執行緒負責。所以它的執行緒模型就是一個消費者一個執行緒。

public class Disruptor<T>
{
    //...
    public Disruptor(final EventFactory<T> eventFactory, final int ringBufferSize, final ThreadFactory threadFactory)
    {
        this(RingBuffer.createMultiProducer(eventFactory, ringBufferSize), new BasicExecutor(threadFactory));
    }
    //...
}

public class BasicExecutor implements Executor
{
    private final ThreadFactory factory;
    private final Queue<Thread> threads = new ConcurrentLinkedQueue<>();

    public BasicExecutor(ThreadFactory factory)
    {
        this.factory = factory;
    }

    @Override
    public void execute(Runnable command)
    {
        //每提交一個任務就新建一個新的執行緒處理這個任務
        final Thread thread = factory.newThread(command);
        if (null == thread)
        {
            throw new RuntimeException("Failed to create thread to run: " + command);
        }

        thread.start();

        threads.add(thread);
    }
    //...
}

 //TODO  後續整理補充一些圖例,方便