spark-streaming的checkpoint机制源码分析

转发请注明原创地址 http://www.cnblogs.com/dongxiao-yang/p/7994357.html

spark-streaming定时对 DStreamGraph 和 JobScheduler 做 Checkpoint,来记录整个 DStreamGraph 的变化和每个 batch 的 job 的完成情况,Checkpoint 发起的间隔默认的是和 batchDuration 一致；即每次 batch 发起、提交了需要运行的 job 后就做 Checkpoint。另外在 job 完成了更新任务状态的时候再次做一下 Checkpoint。

一 checkpoint生成

job生成

  private def generateJobs(time: Time) {
    // Checkpoint all RDDs marked for checkpointing to ensure their lineages are
    // truncated periodically. Otherwise, we may run into stack overflows (SPARK-6847).
    ssc.sparkContext.setLocalProperty(RDD.CHECKPOINT_ALL_MARKED_ANCESTORS, "true")
    Try {
      jobScheduler.receiverTracker.allocateBlocksToBatch(time) // allocate received blocks to batch
      graph.generateJobs(time) // generate jobs using allocated block
    } match {
      case Success(jobs) =>
        val streamIdToInputInfos = jobScheduler.inputInfoTracker.getInfo(time)
        jobScheduler.submitJobSet(JobSet(time, jobs, streamIdToInputInfos))
      case Failure(e) =>
        jobScheduler.reportError("Error generating jobs for time " + time, e)
        PythonDStream.stopStreamingContextIfPythonProcessIsDead(e)
    }
    eventLoop.post(DoCheckpoint(time, clearCheckpointDataLater = false))
  }

job 完成

  private def clearMetadata(time: Time) {
    ssc.graph.clearMetadata(time)

    // If checkpointing is enabled, then checkpoint,
    // else mark batch to be fully processed
    if (shouldCheckpoint) {
      eventLoop.post(DoCheckpoint(time, clearCheckpointDataLater = true))
    } else {
      // If checkpointing is not enabled, then delete metadata information about
      // received blocks (block data not saved in any case). Otherwise, wait for
      // checkpointing of this batch to complete.
      val maxRememberDuration = graph.getMaxInputStreamRememberDuration()
      jobScheduler.receiverTracker.cleanupOldBlocksAndBatches(time - maxRememberDuration)
      jobScheduler.inputInfoTracker.cleanup(time - maxRememberDuration)
      markBatchFullyProcessed(time)
    }
  }

上文里面的eventLoop是JobGenerator内部的一个消息事件队列的封装，eventLoop内部会有一个后台线程不断的去消费事件，所以DoCheckpoint这种类型的事件会经过processEvent ->

doCheckpoint  由checkpointWriter把生成的Checkpoint对象写到外部存储：

  /** Processes all events */
  private def processEvent(event: JobGeneratorEvent) {
    logDebug("Got event " + event)
    event match {
      case GenerateJobs(time) => generateJobs(time)
      case ClearMetadata(time) => clearMetadata(time)
      case DoCheckpoint(time, clearCheckpointDataLater) =>
        doCheckpoint(time, clearCheckpointDataLater)
      case ClearCheckpointData(time) => clearCheckpointData(time)
    }
  }

  /** Perform checkpoint for the give `time`. */
  private def doCheckpoint(time: Time, clearCheckpointDataLater: Boolean) {
    if (shouldCheckpoint && (time - graph.zeroTime).isMultipleOf(ssc.checkpointDuration)) {
      logInfo("Checkpointing graph for time " + time)
      ssc.graph.updateCheckpointData(time)
      checkpointWriter.write(new Checkpoint(ssc, time), clearCheckpointDataLater)
    }
  }

doCheckpoint在调用checkpointWriter写数据到hdfs之前，首先会运行一下ssc.graph.updateCheckpointData(time),这个方法的主要作用是更新DStreamGraph里所有input和output stream对应的checkpointData属性，调用链路为DStreamGraph.updateCheckpointData -> Dstream.updateCheckpointData -> checkpointData.update

  def updateCheckpointData(time: Time) {
    logInfo("Updating checkpoint data for time " + time)
    this.synchronized {
      outputStreams.foreach(_.updateCheckpointData(time))
    }
    logInfo("Updated checkpoint data for time " + time)
  }

  private[streaming] def updateCheckpointData(currentTime: Time) {
    logDebug(s"Updating checkpoint data for time $currentTime")
    checkpointData.update(currentTime)
    dependencies.foreach(_.updateCheckpointData(currentTime))
    logDebug(s"Updated checkpoint data for time $currentTime: $checkpointData")
  }
  private[streaming]
  class DirectKafkaInputDStreamCheckpointData extends DStreamCheckpointData(this) {
    def batchForTime: mutable.HashMap[Time, Array[(String, Int, Long, Long)]] = {
      data.asInstanceOf[mutable.HashMap[Time, Array[OffsetRange.OffsetRangeTuple]]]
    }

    override def update(time: Time): Unit = {
      batchForTime.clear()
      generatedRDDs.foreach { kv =>
        val a = kv._2.asInstanceOf[KafkaRDD[K, V]].offsetRanges.map(_.toTuple).toArray
        batchForTime += kv._1 -> a
      }
    }

    override def cleanup(time: Time): Unit = { }

    override def restore(): Unit = {
      batchForTime.toSeq.sortBy(_._1)(Time.ordering).foreach { case (t, b) =>
         logInfo(s"Restoring KafkaRDD for time $t ${b.mkString("[", ", ", "]")}")
         generatedRDDs += t -> new KafkaRDD[K, V](
           context.sparkContext,
           executorKafkaParams,
           b.map(OffsetRange(_)),
           getPreferredHosts,
           // during restore, it's possible same partition will be consumed from multiple
           // threads, so dont use cache
           false
         )
      }
    }
  }

以DirectKafkaInputDStream为例，代码里重写了checkpointData的update等接口，所以DirectKafkaInputDStream会在checkpoint之前把正在运行的kafkaRDD对应的topic，partition，fromoffset，untiloffset全部存储到checkpointData里面data这个HashMap的属性当中，用于写checkpoint时进行序列化。

一个checkpoint里面包含的对象列表如下：

class Checkpoint(ssc: StreamingContext, val checkpointTime: Time)
  extends Logging with Serializable {
  val master = ssc.sc.master
  val framework = ssc.sc.appName
  val jars = ssc.sc.jars
  val graph = ssc.graph
  val checkpointDir = ssc.checkpointDir
  val checkpointDuration = ssc.checkpointDuration
  val pendingTimes = ssc.scheduler.getPendingTimes().toArray
  val sparkConfPairs = ssc.conf.getAll

二 从checkpoint恢复服务

spark-streaming启用checkpoint代码里的StreamingContext必须严格按照官方demo实例的架构使用，即所有的streaming逻辑都放在一个返回StreamingContext的createContext方法上，

通过StreamingContext.getOrCreate方法进行初始化，在CheckpointReader.read找到checkpoint文件并且成功反序列化出checkpoint对象后，返回一个包含该checkpoint对象的StreamingContext，这个时候程序里的createContext就不会被调用，反之如果程序是第一次启动会通过createContext初始化StreamingContext

  def getOrCreate(
      checkpointPath: String,
      creatingFunc: () => StreamingContext,
      hadoopConf: Configuration = SparkHadoopUtil.get.conf,
      createOnError: Boolean = false
    ): StreamingContext = {
    val checkpointOption = CheckpointReader.read(
      checkpointPath, new SparkConf(), hadoopConf, createOnError)
    checkpointOption.map(new StreamingContext(null, _, null)).getOrElse(creatingFunc())
  }

  def read(
      checkpointDir: String,
      conf: SparkConf,
      hadoopConf: Configuration,
      ignoreReadError: Boolean = false): Option[Checkpoint] = {
    val checkpointPath = new Path(checkpointDir)

    val fs = checkpointPath.getFileSystem(hadoopConf)

    // Try to find the checkpoint files
    val checkpointFiles = Checkpoint.getCheckpointFiles(checkpointDir, Some(fs)).reverse
    if (checkpointFiles.isEmpty) {
      return None
    }

    // Try to read the checkpoint files in the order
    logInfo(s"Checkpoint files found: ${checkpointFiles.mkString(",")}")
    var readError: Exception = null
    checkpointFiles.foreach { file =>
      logInfo(s"Attempting to load checkpoint from file $file")
      try {
        val fis = fs.open(file)
        val cp = Checkpoint.deserialize(fis, conf)
        logInfo(s"Checkpoint successfully loaded from file $file")
        logInfo(s"Checkpoint was generated at time ${cp.checkpointTime}")
        return Some(cp)
      } catch {
        case e: Exception =>
          readError = e
          logWarning(s"Error reading checkpoint from file $file", e)
      }
    }

    // If none of checkpoint files could be read, then throw exception
    if (!ignoreReadError) {
      throw new SparkException(
        s"Failed to read checkpoint from directory $checkpointPath", readError)
    }
    None
  }
}

在从checkpoint恢复的过程中DStreamGraph由checkpoint恢复，下文的代码调用路径StreamingContext.graph->DStreamGraph.restoreCheckpointData ->   DStream.restoreCheckpointData->checkpointData.restore

  private[streaming] val graph: DStreamGraph = {
    if (isCheckpointPresent) {
      _cp.graph.setContext(this)
      _cp.graph.restoreCheckpointData()
      _cp.graph
    } else {
      require(_batchDur != null, "Batch duration for StreamingContext cannot be null")
      val newGraph = new DStreamGraph()
      newGraph.setBatchDuration(_batchDur)
      newGraph
    }
  }

  def restoreCheckpointData() {
    logInfo("Restoring checkpoint data")
    this.synchronized {
      outputStreams.foreach(_.restoreCheckpointData())
    }
    logInfo("Restored checkpoint data")
  }

  private[streaming] def restoreCheckpointData() {
    if (!restoredFromCheckpointData) {
      // Create RDDs from the checkpoint data
      logInfo("Restoring checkpoint data")
      checkpointData.restore()
      dependencies.foreach(_.restoreCheckpointData())
      restoredFromCheckpointData = true
      logInfo("Restored checkpoint data")
    }
  }

    override def restore(): Unit = {
      batchForTime.toSeq.sortBy(_._1)(Time.ordering).foreach { case (t, b) =>
         logInfo(s"Restoring KafkaRDD for time $t ${b.mkString("[", ", ", "]")}")
         generatedRDDs += t -> new KafkaRDD[K, V](
           context.sparkContext,
           executorKafkaParams,
           b.map(OffsetRange(_)),
           getPreferredHosts,
           // during restore, it's possible same partition will be consumed from multiple
           // threads, so dont use cache
           false
         )
      }
    }

仍然以DirectKafkaInputDStreamCheckpointData为例，这个方法从上文保存的checkpoint.data对象里获取RDD运行时的对应信息恢复出停止时的KafkaRDD。

  private def restart() {
    // If manual clock is being used for testing, then
    // either set the manual clock to the last checkpointed time,
    // or if the property is defined set it to that time
    if (clock.isInstanceOf[ManualClock]) {
      val lastTime = ssc.initialCheckpoint.checkpointTime.milliseconds
      val jumpTime = ssc.sc.conf.getLong("spark.streaming.manualClock.jump", 0)
      clock.asInstanceOf[ManualClock].setTime(lastTime + jumpTime)
    }

    val batchDuration = ssc.graph.batchDuration

    // Batches when the master was down, that is,
    // between the checkpoint and current restart time
    val checkpointTime = ssc.initialCheckpoint.checkpointTime
    val restartTime = new Time(timer.getRestartTime(graph.zeroTime.milliseconds))
    val downTimes = checkpointTime.until(restartTime, batchDuration)
    logInfo("Batches during down time (" + downTimes.size + " batches): "
      + downTimes.mkString(", "))

    // Batches that were unprocessed before failure
    val pendingTimes = ssc.initialCheckpoint.pendingTimes.sorted(Time.ordering)
    logInfo("Batches pending processing (" + pendingTimes.length + " batches): " +
      pendingTimes.mkString(", "))
    // Reschedule jobs for these times
    val timesToReschedule = (pendingTimes ++ downTimes).filter { _ < restartTime }
      .distinct.sorted(Time.ordering)
    logInfo("Batches to reschedule (" + timesToReschedule.length + " batches): " +
      timesToReschedule.mkString(", "))
    timesToReschedule.foreach { time =>
      // Allocate the related blocks when recovering from failure, because some blocks that were
      // added but not allocated, are dangling in the queue after recovering, we have to allocate
      // those blocks to the next batch, which is the batch they were supposed to go.
      jobScheduler.receiverTracker.allocateBlocksToBatch(time) // allocate received blocks to batch
      jobScheduler.submitJobSet(JobSet(time, graph.generateJobs(time)))
    }

    // Restart the timer
    timer.start(restartTime.milliseconds)
    logInfo("Restarted JobGenerator at " + restartTime)
  }

最后，在restart的过程中，JobGenerator通过当前时间和上次程序停止的时间计算出程序重启过程中共有多少batch没有生成，加上上一次程序死掉的过程中有多少正在运行的job，全部

进行Reschedule，补跑任务。

参考文档

1Driver 端长时容错详解

2Spark Streaming揭秘 Day33 checkpoint的使用