Flink-checkPoint机制

转载：https://yijiyong.com/dp/flink/05-checkpoints.html

• Flink CheckPoint详细过程
  • 第一步
  • 第二步
  • 第三步
  • 第四步
• 【----------------------------】
• Flink Checkpoint机制分析
• 1. 为什么需要Checkpoint
• 2. Checkpoint执行流程
• 3. Checkpoint相关配置
• 4. Checkpoint过程源码分析
  • 4.1 Client端生成Checkpoint配置
  • 4.2 JobManager发起Checkpoint
  • 4.3 TaskManager执行Checkpoint
  • 4.4 JobManager确认Checkpoint
  • 参考文章

#Flink CheckPoint详细过程

Checkpoint由JM的Checkpoint Coordinator发起

#第一步

Checkpoint Coordinator 向所有 source 节点 trigger Checkpoint；。

wxmp

#第二步

source 节点向下游广播 barrier，这个 barrier 就是实现 Chandy-Lamport 分布式快照算法的核心，下游的 task 只有收到所有 input 的 barrier 才会执行相应的 Checkpoint。

wxmp

#第三步

当 task 完成 state 备份后，会将备份数据的地址（state handle）通知给 Checkpoint coordinator。

wxmp

这里分为同步和异步（如果开启的话）两个阶段：

1.同步阶段：task执行状态快照，并写入外部存储系统（根据状态后端的选择不同有所区别）

执行快照的过程：

• a.对state做深拷贝。
• b.将写操作封装在异步的FutureTask中

FutureTask的作用包括：

• 1）打开输入流
• 2）写入状态的元数据信息
• 3）写入状态
• 4）关闭输入流

2.异步阶段：

• 1）执行同步阶段创建的FutureTask
• 2）向Checkpoint Coordinator发送ACK响应

#第四步

下游的 sink 节点收集齐上游两个 input 的 barrier 之后，会执行本地快照，这里特地展示了 RocksDB incremental Checkpoint 的流程，首先 RocksDB 会全量刷数据到磁盘上（红色大三角表示），然后 Flink 框架会从中选择没有上传的文件进行持久化备份（紫色小三角）。

wxmp

同样的，sink 节点在完成自己的 Checkpoint 之后，会将 state handle 返回通知 Coordinator。

wxmp

最后，当 Checkpoint coordinator 收集齐所有 task 的 state handle，就认为这一次的 Checkpoint 全局完成了，向持久化存储中再备份一个 Checkpoint meta 文件。

wxmp

#【----------------------------】

#Flink Checkpoint机制分析

原创作品，转载请标明 ：http://blog.csdn.net/xiejingfa/article/details/105439802

可靠性是分布式系统实现必须考虑的因素之一。Flink基于Chandy-Lamport分布式快照算法实现了一套可靠的Checkpoint机制，可以保证集群中某些节点出现故障时，能够将整个作业恢复到故障之前某个状态。同时，Checkpoint机制也是Flink实现Exactly-Once语义的基础。

本文将介绍Flink的Checkpoint机制的原理，并从源码层面了解Checkpoint机制是如何实现的（基于Flink 1.10）。

#1. 为什么需要Checkpoint

Flink是有状态的流计算处理引擎，每个算子Operator可能都需要记录自己的运行数据，并在接收到新流入的元素后不断更新自己的状态数据。当分布式系统引入状态计算后，为了保证计算结果的正确性（特别是对于流处理系统，不可能每次系统故障后都从头开始计算），就必然要求系统具有容错性。对于Flink来说，Flink作业运行在多个节点上，当出现节点宕机、网络故障等问题，需要一个机制保证节点保存在本地的状态不丢失。流处理中Exactly-Once语义的实现也要求作业从失败恢复后的状态要和失败前的状态一致。

那么怎么保证分布式环境下各节点状态的容错呢？通常这是通过定期对作业状态和数据流进行快照实现的，常见的检查点算法有比如Sync-and-Stop（SNS）算法、Chandy-Lamport（CL）算法。

Flink的Checkpoint机制是基于Chandy-Lamport算法的思想改进而来，引入了Checkpoint Barrier的概念，可以在不停止整个流处理系统的前提下，让每个节点独立建立检查点保存自身快照，并最终达到整个作业全局快照的状态。有了全局快照，当我们遇到故障或者重启的时候就可以直接从快照中恢复，这就是Flink容错的核心。

#2. Checkpoint执行流程

Barrier是Flink分布式快照的核心概念之一，称之为屏障或者数据栅栏（可以理解为快照的分界线）。Barrier是一种特殊的内部消息，在进行Checkpoint的时候Flink会在数据流源头处周期性地注入Barrier，这些Barrier会作为数据流的一部分，一起流向下游节点并且不影响正常的数据流。Barrier的作用是将无界数据流从时间上切分成多个窗口，每个窗口对应一系列连续的快照中的一个，每个Barrier都带有一个快照ID，一个Barrier生成之后，在这之前的数据都进入此快照，在这之后的数据则进入下一个快照。

wxmp

如上图，Barrier-n跟随着数据流一起流动，当算子从输入流接收到Barrier-n后，就会停止接收数据并对当前自身的状态做一次快照，快照完成后再将Barrier-n以广播的形式传给下游节点。一旦作业的Sink算子接收到Barrier n后，会向JobMnager发送一个消息，确认Barrier-n对应的快照完成。当作业中的所有Sink算子都确认后，意味一次全局快照也就完成。

当一个算子有多个上游节点时，会接收到多个Barrier，这时候需要进行Barrier Align对齐操作。

wxmp

如上图，一个算子有两个输入流，当算子从一个上游数据流接收到一个Barrier-n后，它不会立即向下游广播，而是先暂停对该数据流的处理，将到达的数据先缓存在Input Buffer中（因为这些数据属于下一次快照而不是当前快照，缓存数据可以不阻塞该数据流），直到从另外一个数据流中接收到Barrier-n，才会进行快照处理并将Barrier-n向下游发送。从这个流程可以看出，如果开启Barrier对齐后，算子由于需要等待所有输入节点的Barrier到来出现暂停，对整体的性能也会有一定的影响。

综上，Flink Checkpoint机制的核心思想实质上是通过Barrier来标记触发快照的时间点和对应需要进行快照的数据集，将数据流处理和快照操作解耦开来，从而最大程度降低快照对系统性能的影响。

Flink的一致性和Checkpoint机制有紧密的关系：

• 当不开启Checkpoint时，节点发生故障时可能会导致数据丢失，这就是At-Most-Once
• 当开启Checkpoint但不进行Barrier对齐时，对于有多个输入流的节点如果发生故障，会导致有一部分数据可能会被处理多次，这就是At-Least-Once
• 当开启Checkpoint并进行Barrier对齐时，可以保证每条数据在故障恢复时只会被重放一次，这就是Exactly-Once

#3. Checkpoint相关配置

StreamExecutionEnvironment env = StreamExecutionEnvironment.getExecutionEnvironment();
env.enableCheckpointing(1000);
env.getCheckpointConfig().setCheckpointingMode(CheckpointingMode.EXACTLY_ONCE);
env.getCheckpointConfig().setMinPauseBetweenCheckpoints(500);
env.getCheckpointConfig().setCheckpointTimeout(60000);
env.getCheckpointConfig().setMaxConcurrentCheckpoints(1);
env.getCheckpointConfig().enableExternalizedCheckpoints(ExternalizedCheckpointCleanup.RETAIN_ON_CANCELLATION);

• 默认情况下，Checkpoint机制是关闭的，需要通过enableCheckpointing(interval)来开启，并指定每interval毫秒进行一次Checkpoint。
• Checkpoint模式支持Exactly-Once和At-Least-Once，可以通过setCheckpointingMode来设置。
• 如果两次Checkpoint的时间很短，会导致整个系统大部分资源都用于执行Checkpoint，影响正常作业的执行。可以通过setMinPauseBetweenCheckpoints来设置两次Checkpoint之间的最小间隔。
• setCheckpointTimeout可以给Checkpoint设置一个超时时间，当一次Checkpoint超过一定时间没有完成，直接终止掉。
• 默认情况下，当一个Checkpoint还在执行时，不会触发另一个Checkpoint，通过setMaxConcurrentCheckpoints可以设置最大并发Checkpoint数量。
• enableExternalizedCheckpoints可以设置当用户取消了作业后，是否保留远程存储上的Checkpoint数据，一般设置为RETAIN_ON_CANCELLATION。

#4. Checkpoint过程源码分析

#4.1 Client端生成Checkpoint配置

Client端在向JobManger提交作业前会根据用户代码生成StreamGraph，再转化为JobGraph，在构建JobGraph时会调用configureCheckpointing生成JobCheckpointingSettings配置并保存在JobGraph中。这里要注意到triggerVertices这个集合，它表示Flink通过这些节点触发Checkpoint。在构建JobGraph时只会将Source节点加入到triggerVertices，决定Flink之后发起Checkpoint时只针对Source节点注入Barrier。

private void configureCheckpointing() {
    CheckpointConfig cfg = streamGraph.getCheckpointConfig();
    ...（省略部分代码，只展示核心流程，下同）

    //  --- configure the participating vertices ---

    // collect the vertices that receive "trigger checkpoint" messages.
    // currently, these are all the sources
    List<JobVertexID> triggerVertices = new ArrayList<>();

    // collect the vertices that need to acknowledge the checkpoint
    // currently, these are all vertices
    List<JobVertexID> ackVertices = new ArrayList<>(jobVertices.size());

    // collect the vertices that receive "commit checkpoint" messages
    // currently, these are all vertices
    List<JobVertexID> commitVertices = new ArrayList<>(jobVertices.size());

    for (JobVertex vertex : jobVertices.values()) {
        // 这里只会将Source节点加入到triggerVertices
        if (vertex.isInputVertex()) {  
            triggerVertices.add(vertex.getID());
        }
        commitVertices.add(vertex.getID());
        ackVertices.add(vertex.getID());
    }

    // settings将所有Checkpoint配置封装在一起
    jobGraph.setSnapshotSettings(settings);  
}

#4.2 JobManager发起Checkpoint

CheckpointCoordinator是Flink执行Checkpoint的核心组件，JobManager在接收到Client端的SubmitJob请求后将JobGraph转化为ExecutionGraph，并调用enableCheckpointing方法初始化CheckpointCoordinator，然后为CheckpointCoordinator注册一个Job状态变化的监听器CheckpointCoordinatorDeActivator。

public void enableCheckpointing() {
    ... 

    // create the coordinator that triggers and commits checkpoints and holds the state
    checkpointCoordinator = new CheckpointCoordinator(
    jobInformation.getJobId(),
        chkConfig,
        tasksToTrigger,
        tasksToWaitFor,
        tasksToCommitTo,
        checkpointIDCounter,
        checkpointStore,
        checkpointStateBackend,
        ioExecutor,
        new ScheduledExecutorServiceAdapter(checkpointCoordinatorTimer),
        SharedStateRegistry.DEFAULT_FACTORY,
        failureManager);

    if (chkConfig.getCheckpointInterval() != Long.MAX_VALUE) {
        // the periodic checkpoint scheduler is activated and deactivated as a result of
        // job status changes (running -> on, all other states -> off)
        registerJobStatusListener(checkpointCoordinator.createActivatorDeactivator());
    }
}

CheckpointCoordinatorDeActivator实现了JobStatusListener接口，当job状态变成Running时，调用startCheckpointScheduler方法开启CheckpointScheduler，当job变成其他状态时，调用stopCheckpointScheduler方法停止CheckpointScheduler。

public class CheckpointCoordinatorDeActivator implements JobStatusListener {

    private final CheckpointCoordinator coordinator;

    public CheckpointCoordinatorDeActivator(CheckpointCoordinator coordinator) {
        this.coordinator = checkNotNull(coordinator);
    }

    @Override
    public void jobStatusChanges(JobID jobId, JobStatus newJobStatus, long timestamp, Throwable error) {
        if (newJobStatus == JobStatus.RUNNING) {
            // start the checkpoint scheduler
            coordinator.startCheckpointScheduler();
        } else {
            // anything else should stop the trigger for now
            coordinator.stopCheckpointScheduler();
        }
    }
}

接下来我们来看下startCheckpointScheduler，startCheckpointScheduler首先调用stopCheckpointScheduler确保之前的Checkpoint Scheduler已经停止，然后再创建一个新的ScheduledTrigger放到线程池中定时执行triggerCheckpoint方法触发Checkpoint。第3小节中提到的enableCheckpointing(interval)方法可以设置Checkpoint执行的时间间隔，背后的原理就在这里。

public void startCheckpointScheduler() {
    synchronized (lock) {
        if (shutdown) {
            throw new IllegalArgumentException("Checkpoint coordinator is shut down");
        }

        // make sure all prior timers are cancelled
        stopCheckpointScheduler();

        periodicScheduling = true;
        currentPeriodicTrigger = scheduleTriggerWithDelay(getRandomInitDelay());
    }
}

private ScheduledFuture<?> scheduleTriggerWithDelay(long initDelay) {
    return timer.scheduleAtFixedRate(
        new ScheduledTrigger(),
            initDelay, baseInterval, TimeUnit.MILLISECONDS);
}

private final class ScheduledTrigger implements Runnable {

    @Override
    public void run() {
        try {
            triggerCheckpoint(System.currentTimeMillis(), true);
        }
        catch (Exception e) {
            LOG.error("Exception while triggering checkpoint for job {}.", job, e);
        }
    }
}

triggerCheckpoint是触发Checkpoint的核心方法，下面介绍一些它主要做了哪些工作。

• 检查当前正在处理的并发Checkpoint数是否超过阈值和距离上一次Checkpoint是否小于设置的最小间隔。如果条件不满足，直接返回。

// preCheckBeforeTriggeringCheckpoint是在triggerCheckpoint中调用的方法
private void preCheckBeforeTriggeringCheckpoint(boolean isPeriodic, boolean forceCheckpoint) throws CheckpointException {
    // abort if the coordinator has been shutdown in the meantime
    if (shutdown) {
        throw new CheckpointException(CheckpointFailureReason.CHECKPOINT_COORDINATOR_SHUTDOWN);
    }

    // Don't allow periodic checkpoint if scheduling has been disabled
    if (isPeriodic && !periodicScheduling) {
        throw new CheckpointException(CheckpointFailureReason.PERIODIC_SCHEDULER_SHUTDOWN);
    }

    if (!forceCheckpoint) {
        if (triggerRequestQueued) {
            throw new CheckpointException(CheckpointFailureReason.ALREADY_QUEUED);
        }

        checkConcurrentCheckpoints();

        checkMinPauseBetweenCheckpoints();
    }
}

• 检查所有需要被trigger和ack的Task是否都处于运行状态，只要有一个Task不满足条件，就没有必要触发本次Checkpoint了。

// check if all tasks that we need to trigger are running.
// if not, abort the checkpoint
Execution[] executions = new Execution[tasksToTrigger.length];
for (int i = 0; i < tasksToTrigger.length; i++) {
    Execution ee = tasksToTrigger[i].getCurrentExecutionAttempt();
    if (ee == null) {
        throw new CheckpointException(CheckpointFailureReason.NOT_ALL_REQUIRED_TASKS_RUNNING);
    } else if (ee.getState() == ExecutionState.RUNNING) {
        executions[i] = ee;
    } else {
        throw new CheckpointException(CheckpointFailureReason.NOT_ALL_REQUIRED_TASKS_RUNNING);
    }
}

// next, check if all tasks that need to acknowledge the checkpoint are running.
// if not, abort the checkpoint
Map<ExecutionAttemptID, ExecutionVertex> ackTasks = new HashMap<>(tasksToWaitFor.length);
for (ExecutionVertex ev : tasksToWaitFor) {
    Execution ee = ev.getCurrentExecutionAttempt();
    if (ee != null) {
        ackTasks.put(ee.getAttemptId(), ev);
    } else {
        throw new CheckpointException(CheckpointFailureReason.NOT_ALL_REQUIRED_TASKS_RUNNING);
    }
}

• 只有上面两次检查都通过后，才会正在进入Checkpoint的处理流程。首先生成一个新的CheckpointID，再创建一个PendingCheckpoint对象。PendingCheckpoint是一个启动但还未被确认的Checkpoint。等到所有Task都确认后又会转化为CompletedCheckpoint。

// we will actually trigger this checkpoint!
final CheckpointStorageLocation checkpointStorageLocation;
final long checkpointID;

try {
    // this must happen outside the coordinator-wide lock, because it communicates
    // with external services (in HA mode) and may block for a while.
    checkpointID = checkpointIdCounter.getAndIncrement();
}
catch (Throwable t) {
    ...
}

final PendingCheckpoint checkpoint = new PendingCheckpoint(
    job,
    checkpointID,
    timestamp,
    ackTasks,
    masterHooks.keySet(),
    props,
    checkpointStorageLocation,
    executor);

• 为了防止Checkpoint长时间未完成而占用资源，CheckpointCoordinator还会创建一个取消器用于清理超时的Checkpoint。

// schedule the timer that will clean up the expired checkpoints
final Runnable canceller = () -> {
    synchronized (lock) {
        // only do the work if the checkpoint is not discarded anyways
        // note that checkpoint completion discards the pending checkpoint object
        if (!checkpoint.isDiscarded()) {
            failPendingCheckpoint(checkpoint, CheckpointFailureReason.CHECKPOINT_EXPIRED);
            pendingCheckpoints.remove(checkpointID);
            rememberRecentCheckpointId(checkpointID);

            triggerQueuedRequests();
        }
    }
};

ScheduledFuture<?> cancellerHandle = timer.schedule(canceller, checkpointTimeout, TimeUnit.MILLISECONDS);

• 最后向Source节点发送消息，触发Checkpoint。

// send the messages to the tasks that trigger their checkpoint
for (Execution execution: executions) {
    if (props.isSynchronous()) {
        execution.triggerSynchronousSavepoint(checkpointID, timestamp, checkpointOptions, advanceToEndOfTime);
    } else {
        execution.triggerCheckpoint(checkpointID, timestamp, checkpointOptions);
    }
}

#4.3 TaskManager执行Checkpoint

TaskManager执行Checkpoint操作要分以下两种情况来讨论：

• Source节点接收到JobManager发送的TriggerCheckpoint消息后触发本节点Checkpoint。
• 非Source节点从上游接收到Barrier后触发本节点Checkpoint，这里可能还会涉及到Barrier对齐操作。

#4.3.1 Source节点执行Checkpoint

下面先来看看Source节点是如何执行Checkpoint的。

TaskManager接收到JobManager的TriggerCheckpoint消息后，经过层层调用最后使用AbstractInvokable的triggerCheckpointAsync方法来处理。AbstractInvokable是对在TaskManager中可执行任务的抽象。triggerCheckpointAsync的具体实现在AbstractInvokable的子类StreamTask中，其核心逻辑就是使用线程池异步调用triggerCheckpoint方法。

public Future<Boolean> triggerCheckpointAsync(
        CheckpointMetaData checkpointMetaData,
        CheckpointOptions checkpointOptions,
        boolean advanceToEndOfEventTime) {

    return mailboxProcessor.getMainMailboxExecutor().submit(
            () -> triggerCheckpoint(checkpointMetaData, checkpointOptions, advanceToEndOfEventTime),
            "checkpoint %s with %s",
        checkpointMetaData,
        checkpointOptions);
}

private boolean triggerCheckpoint(
        CheckpointMetaData checkpointMetaData,
        CheckpointOptions checkpointOptions,
        boolean advanceToEndOfEventTime) throws Exception {
    try {
        ...

        boolean success = performCheckpoint(checkpointMetaData, checkpointOptions, checkpointMetrics, advanceToEndOfEventTime);
        if (!success) {
            declineCheckpoint(checkpointMetaData.getCheckpointId());
        }
        return success;
    } catch (Exception e) {
        ...
    }
}

StreamTask的triggerCheckpoint会调用performCheckpoint方法，该方法主要工作包括：

• 创建Checkpoint Barrier并向下游节点广播。
• 触发本节点的快照操作。

private boolean performCheckpoint(
        CheckpointMetaData checkpointMetaData,
        CheckpointOptions checkpointOptions,
        CheckpointMetrics checkpointMetrics,
        boolean advanceToEndOfTime) throws Exception {
    ...

    final long checkpointId = checkpointMetaData.getCheckpointId();

    if (isRunning) {
        actionExecutor.runThrowing(() -> {
            ...
            // All of the following steps happen as an atomic step from the perspective of barriers and
            // records/watermarks/timers/callbacks.
            // We generally try to emit the checkpoint barrier as soon as possible to not affect downstream
            // checkpoint alignments

            // Step (1): Prepare the checkpoint, allow operators to do some pre-barrier work.
            //           The pre-barrier work should be nothing or minimal in the common case.
            operatorChain.prepareSnapshotPreBarrier(checkpointId);

            // Step (2): Send the checkpoint barrier downstream
            operatorChain.broadcastCheckpointBarrier(
                    checkpointId,
                    checkpointMetaData.getTimestamp(),
                    checkpointOptions);

            // Step (3): Take the state snapshot. This should be largely asynchronous, to not
            //           impact progress of the streaming topology
            checkpointState(checkpointMetaData, checkpointOptions, checkpointMetrics);
        });

        return true;
    } else {
        ...
        return false;
    }
}

checkpointState方法进一步调用executeCheckpointing对本地的State进行保存，该方法被封装在CheckpointingOperation类中，其核心工作包括：

• 调用每一个StreaOperator的snapshotState方法生成快照并存储到状态后端。
• 检查Checkpoint结果并告诉JobManager。

public void executeCheckpointing() throws Exception {
    startSyncPartNano = System.nanoTime();

    try {
        // 调用每一个算子的snapshotState方法
        for (StreamOperator<?> op : allOperators) {
            checkpointStreamOperator(op);
        }

        startAsyncPartNano = System.nanoTime();

        checkpointMetrics.setSyncDurationMillis((startAsyncPartNano - startSyncPartNano) / 1_000_000);

        // we are transferring ownership over snapshotInProgressList for cleanup to the thread, active on submit
        AsyncCheckpointRunnable asyncCheckpointRunnable = new AsyncCheckpointRunnable(
            owner,
            operatorSnapshotsInProgress,
            checkpointMetaData,
            checkpointMetrics,
            startAsyncPartNano);

        owner.cancelables.registerCloseable(asyncCheckpointRunnable);
        // 检查结果并报告JobManager
        owner.asyncOperationsThreadPool.execute(asyncCheckpointRunnable);
    } catch (Exception ex) {
        ...
    }
}

private void checkpointStreamOperator(StreamOperator<?> op) throws Exception {
    if (null != op) {
        OperatorSnapshotFutures snapshotInProgress = op.snapshotState(
                checkpointMetaData.getCheckpointId(),
                checkpointMetaData.getTimestamp(),
                checkpointOptions,
                storageLocation);
        operatorSnapshotsInProgress.put(op.getOperatorID(), snapshotInProgress);
    }
}

如果Checkpoint执行成功，AsyncCheckpointRunnable最后会调用TaskStateManagerImpl的reportTaskStateSnapshots方法向JobManager发送AcknowledgeCheckpoint消息。

public void reportTaskStateSnapshots(
        @Nonnull CheckpointMetaData checkpointMetaData,
        @Nonnull CheckpointMetrics checkpointMetrics,
        @Nullable TaskStateSnapshot acknowledgedState,
        @Nullable TaskStateSnapshot localState) {

    long checkpointId = checkpointMetaData.getCheckpointId();

    localStateStore.storeLocalState(checkpointId, localState);

    checkpointResponder.acknowledgeCheckpoint(
        jobId,
        executionAttemptID,
        checkpointId,
        checkpointMetrics,
        acknowledgedState);
}

#4.3.1 非Source节点执行Checkpoint

下游的非Source节点接收到Barrier后，调用CheckpointBarrierAligner的processBarrier方法来处理。processBarrier会分别处理单个Input Channel和多个Input Channel两个不同场景，具体为：

• 如果只有一个Input Channel，收到Barrier后直接调用notifyCheckpoint触发快照。
• 如果包含多个Input Channel， 先执行Barrier对齐，收到所有Input Channel发送的Barrier后再调用notifyCheckpoint触发快照。

public boolean processBarrier(CheckpointBarrier receivedBarrier, int channelIndex, long bufferedBytes) throws Exception {
    final long barrierId = receivedBarrier.getId();

    // fast path for single channel cases
    if (totalNumberOfInputChannels == 1) {
        if (barrierId > currentCheckpointId) {
            // new checkpoint
            currentCheckpointId = barrierId;
            notifyCheckpoint(receivedBarrier, bufferedBytes, latestAlignmentDurationNanos);
        }
        return false;
    }

    boolean checkpointAborted = false;

    // -- general code path for multiple input channels --

    if (numBarriersReceived > 0) {
        // this is only true if some alignment is already progress and was not canceled

        if (barrierId == currentCheckpointId) {
            // regular case
            onBarrier(channelIndex);
        }
        else if (barrierId > currentCheckpointId) {
            ...

            // abort the current checkpoint
            releaseBlocksAndResetBarriers();
            checkpointAborted = true;

            // begin a the new checkpoint
            beginNewAlignment(barrierId, channelIndex);
        }
        else {
            // ignore trailing barrier from an earlier checkpoint (obsolete now)
            return false;
        }
    }
    else if (barrierId > currentCheckpointId) {
        // first barrier of a new checkpoint
        beginNewAlignment(barrierId, channelIndex);
    }
    else {
        // either the current checkpoint was canceled (numBarriers == 0) or
        // this barrier is from an old subsumed checkpoint
        return false;
    }

    // check if we have all barriers - since canceled checkpoints always have zero barriers
    // this can only happen on a non canceled checkpoint
    if (numBarriersReceived + numClosedChannels == totalNumberOfInputChannels) {
        // actually trigger checkpoint
        releaseBlocksAndResetBarriers();
        notifyCheckpoint(receivedBarrier, bufferedBytes, latestAlignmentDurationNanos);
        return true;
    }
    return checkpointAborted;
}

toNotifyOnCheckpoint是AbstractInvokable实例，triggerCheckpointOnBarrier方法最终调用了performCheckpoint方法，这后面的逻辑就跟Source节点一样了。可以看出：Source节点和非Source节点执行快照的逻辑是一致的，不同的是触发快照的机制。Source节点接收到JobManager发送的TriggerCheckpoint消息触发快照，非Source节点接收到上游节点的Barrier后触发快照。

// CheckpointBarrierHandler
protected void notifyCheckpoint(CheckpointBarrier checkpointBarrier, long bufferedBytes, long alignmentDurationNanos) throws Exception {
    if (toNotifyOnCheckpoint != null) {
        CheckpointMetaData checkpointMetaData =
            new CheckpointMetaData(checkpointBarrier.getId(), checkpointBarrier.getTimestamp());
        ...
        toNotifyOnCheckpoint.triggerCheckpointOnBarrier(
            checkpointMetaData,
            checkpointBarrier.getCheckpointOptions(),
            checkpointMetrics);
    }
}

// StreamTask
public void triggerCheckpointOnBarrier(
        CheckpointMetaData checkpointMetaData,
        CheckpointOptions checkpointOptions,
        CheckpointMetrics checkpointMetrics) throws Exception {
    try {
        if (performCheckpoint(checkpointMetaData, checkpointOptions, checkpointMetrics, false)) {
            if (isSynchronousSavepointId(checkpointMetaData.getCheckpointId())) {
                runSynchronousSavepointMailboxLoop();
            }
        }
    }
    catch (Exception e) {
        ...
    }
}

#4.4 JobManager确认Checkpoint

JobManager收到Task的AcknowledgeCheckpoint消息后，会调用CheckpointCoordinator的receiveAcknowledgeMessage方法来处理。PendingCheckPoint中记录了本次Checkpoint中有哪些Task需要Ack，如果JobManager已经收到所有的Task的Ack消息，则调用completePendingCheckpoint向Task发送notifyCheckpointComplete消息通知Task本次Checkpoint已经完成。

final PendingCheckpoint checkpoint = pendingCheckpoints.get(checkpointId);

if (checkpoint != null && !checkpoint.isDiscarded()) {
    switch (checkpoint.acknowledgeTask(message.getTaskExecutionId(), message.getSubtaskState(), message.getCheckpointMetrics())){
        case SUCCESS:
            if (checkpoint.areTasksFullyAcknowledged()) {
                completePendingCheckpoint(checkpoint);
            }
            break;
        ...
    }
}

private void completePendingCheckpoint(PendingCheckpoint pendingCheckpoint) throws CheckpointException {
    ...

    // send the "notify complete" call to all vertices
    final long timestamp = completedCheckpoint.getTimestamp();

    for (ExecutionVertex ev : tasksToCommitTo) {
        Execution ee = ev.getCurrentExecutionAttempt();
        if (ee != null) {
            ee.notifyCheckpointComplete(checkpointId, timestamp);
        }
    }
}

TaskManager收到notifyCheckpointComplete消息后，最终调用Task的notifyCheckpointComplete方法回调每一个算子的notifyCheckpointComplete方法。

// TaskExecutor
public CompletableFuture<Acknowledge> confirmCheckpoint(
        ExecutionAttemptID executionAttemptID,
        long checkpointId,
        long checkpointTimestamp) {
    final Task task = taskSlotTable.getTask(executionAttemptID);

    if (task != null) {
        task.notifyCheckpointComplete(checkpointId);

        return CompletableFuture.completedFuture(Acknowledge.get());
    } else {
        ...
    }
}

// StreamTask
private void notifyCheckpointComplete(long checkpointId) {
    try {
        boolean success = actionExecutor.call(() -> {
            if (isRunning) {
                for (StreamOperator<?> operator : operatorChain.getAllOperators()) {
                    if (operator != null) {
                        operator.notifyCheckpointComplete(checkpointId);
                    }
                }
                return true;
            } 
            ...
        });
    } catch (Exception e) {
        ...
    }
}

至此，一次完整的Checkpoint过程就完成了。

参考：

1. Lightweight Asynchronous Snapshots for Distributed Dataflows(opens new window)

2. Data Streaming Fault Tolerance(opens new window)

3. Flink Checkpointing(opens new window)

4. Flink 轻量级异步快照ABS实现原理(opens new window)

#参考文章

• https://www.jianshu.com/p/9c587bd491fc
• https://blog.csdn.net/Xiejingfa/article/details/105439802