Spark源码分析：DAGScheduler

概述

在RDD一文中提到：

定义RDD之后，程序员就可以在动作（注：即action操作）中使用RDD了。动作是向应用程序返回值，或向存储系统导出数据的那些操作，例如，count（返回RDD中的元素个数），collect（返回元素本身），save（将RDD输出到存储系统）。在Spark中，只有在动作第一次使用RDD时，才会计算RDD（即延迟计算）。这样在构建RDD的时候，运行时通过管道的方式传输多个转换。

一次action操作会触发RDD的延迟计算，我们把这样的一次计算称作一个Job。我们还提到了窄依赖和宽依赖的概念：

窄依赖指的是：每个parent RDD 的 partition 最多被 child RDD的一个partition使用

宽依赖指的是：每个parent RDD 的 partition 被多个 child RDD的partition使用

窄依赖每个child RDD 的partition的生成操作都是可以并行的，而宽依赖则需要所有的parent partition shuffle结果得到后再进行。

由于在RDD的一系类转换中，若其中一些连续的转换都是窄依赖，那么它们是可以并行的，而有宽依赖则不行。所有，Spark将宽依赖为划分界限，将Job换分为多个Stage。而一个Stage里面的转换任务，我们可以把它抽象成TaskSet。一个TaskSet中有很多个Task，它们的转换操作都是相同的，不同只是操作的对象是对数据集中的不同子数据集。

接下来，Spark就可以提交这些任务了。但是，如何对这些任务进行调度和资源分配呢？如何通知worker去执行这些任务呢？接下来，我们会一一讲解。

根据以上两个阶段，我们会来详细介绍两个Scheduler，一个是DAGScheduler，另外一个是TaskScheduler。

基本概念

• application
• job
• stage
• taskset
• task

DAGScheduler创建

我们先来看一来在SparkContext中是如何创建它们的：

  val (sched, ts) = SparkContext.createTaskScheduler(this, master, deployMode)
    _schedulerBackend = sched
    _taskScheduler = ts
    _dagScheduler = new DAGScheduler(this)  

可以看到，我们是先用函数createTaskScheduler创建了taskScheduler，再new了一个DAGScheduler。这个顺序可以改变吗？答案是否定的，我们看下DAGScheduler类就知道了：

class DAGScheduler(
    private[scheduler] val sc: SparkContext,
    private[scheduler] val taskScheduler: TaskScheduler,
    listenerBus: LiveListenerBus,
    mapOutputTracker: MapOutputTrackerMaster,
    blockManagerMaster: BlockManagerMaster,
    env: SparkEnv,
    clock: Clock = new SystemClock())
  extends Logging {

  def this(sc: SparkContext, taskScheduler: TaskScheduler) = {
    this(
      sc,
      taskScheduler,
      sc.listenerBus,
      sc.env.mapOutputTracker.asInstanceOf[MapOutputTrackerMaster],
      sc.env.blockManager.master,
      sc.env)
  }

  def this(sc: SparkContext) = this(sc, sc.taskScheduler)

***

  }

SparkContext中创建的TaskScheduler，会传入DAGScheduler赋值给它的成员变量，再DAG阶段结束后，使用它进行下一步对任务调度等的操作。

提交job

调用栈如下：

• rdd.count
  • SparkContext.runJob
    • DAGScheduler.runJob
      • DAGScheduler.submitJob
        • DAGSchedulerEventProcessLoop.doOnReceive
          • DAGScheduler.handleJobSubmitted

1.rdd.count

RDD的一些action操作都会触发SparkContext的runJob函数，如count()

/**
 * Return the number of elements in the RDD.
 */
def count(): Long = sc.runJob(this, Utils.getIteratorSize _).sum

2.SparkContext.runJob

SparkContext的runJob会触发 DAGScheduler的runJob：

/**
 * Run a function on a given set of partitions in an RDD and pass the results to the given
 * handler function. This is the main entry point for all actions in Spark.
 *
 * @param rdd target RDD to run tasks on
 * @param func a function to run on each partition of the RDD
 * @param partitions set of partitions to run on; some jobs may not want to compute on all
 * partitions of the target RDD, e.g. for operations like `first()`
 * @param resultHandler callback to pass each result to
 */
def runJob[T, U: ClassTag](
    rdd: RDD[T],
    func: (TaskContext, Iterator[T]) => U,
    partitions: Seq[Int],
    resultHandler: (Int, U) => Unit): Unit = {
  if (stopped.get()) {
    throw new IllegalStateException("SparkContext has been shutdown")
  }
  val callSite = getCallSite
  val cleanedFunc = clean(func)
  logInfo("Starting job: " + callSite.shortForm)
  if (conf.getBoolean("spark.logLineage", false)) {
    logInfo("RDD's recursive dependencies:\n" + rdd.toDebugString)
  }
  dagScheduler.runJob(rdd, cleanedFunc, partitions, callSite, resultHandler, localProperties.get)
  progressBar.foreach(_.finishAll())
  rdd.doCheckpoint()
}

3.dagscheduler.submitjob

  def submitJob[T, U](
      rdd: RDD[T],
      func: (TaskContext, Iterator[T]) => U,
      partitions: Seq[Int],
      callSite: CallSite,
      resultHandler: (Int, U) => Unit,
      properties: Properties): JobWaiter[U] = {
    // 确认没在不存在的partition上执行任务
    val maxPartitions = rdd.partitions.length
    partitions.find(p => p >= maxPartitions || p < 0).foreach { p =>
      throw new IllegalArgumentException(
        "Attempting to access a non-existent partition: " + p + ". " +
          "Total number of partitions: " + maxPartitions)
    }
    //递增得到jobId
    val jobId = nextJobId.getAndIncrement()
    if (partitions.size == 0) {
      //若Job没对任何一个partition执行任务，
      //则立即返回
      return new JobWaiter[U](this, jobId, 0, resultHandler)
    }

    assert(partitions.size > 0)
    val func2 = func.asInstanceOf[(TaskContext, Iterator[_]) => _]
    val waiter = new JobWaiter(this, jobId, partitions.size, resultHandler)
    eventProcessLoop.post(JobSubmitted(
      jobId, rdd, func2, partitions.toArray, callSite, waiter,
      SerializationUtils.clone(properties)))
    waiter
  }

Dagscheduler戳出来你可能不信。其实看看注释就好。==

/**
 * The high-level scheduling layer that implements stage-oriented scheduling. It computes a DAG of
 * stages for each job, keeps track of which RDDs and stage outputs are materialized, and finds a
 * minimal schedule to run the job. It then submits stages as TaskSets to an underlying
 * TaskScheduler implementation that runs them on the cluster. A TaskSet contains fully independent
 * tasks that can run right away based on the data that's already on the cluster (e.g. map output
 * files from previous stages), though it may fail if this data becomes unavailable.
 *
 * Spark stages are created by breaking the RDD graph at shuffle boundaries. RDD operations with
 * "narrow" dependencies, like map() and filter(), are pipelined together into one set of tasks
 * in each stage, but operations with shuffle dependencies require multiple stages (one to write a
 * set of map output files, and another to read those files after a barrier). In the end, every
 * stage will have only shuffle dependencies on other stages, and may compute multiple operations
 * inside it. The actual pipelining of these operations happens in the RDD.compute() functions of
 * various RDDs (MappedRDD, FilteredRDD, etc).
 *
 * In addition to coming up with a DAG of stages, the DAGScheduler also determines the preferred
 * locations to run each task on, based on the current cache status, and passes these to the
 * low-level TaskScheduler. Furthermore, it handles failures due to shuffle output files being
 * lost, in which case old stages may need to be resubmitted. Failures *within* a stage that are
 * not caused by shuffle file loss are handled by the TaskScheduler, which will retry each task
 * a small number of times before cancelling the whole stage.
 *
 * When looking through this code, there are several key concepts:
 *
 *  - Jobs (represented by [[ActiveJob]]) are the top-level work items submitted to the scheduler.
 *    For example, when the user calls an action, like count(), a job will be submitted through
 *    submitJob. Each Job may require the execution of multiple stages to build intermediate data.
 *
 *  - Stages ([[Stage]]) are sets of tasks that compute intermediate results in jobs, where each
 *    task computes the same function on partitions of the same RDD. Stages are separated at shuffle
 *    boundaries, which introduce a barrier (where we must wait for the previous stage to finish to
 *    fetch outputs). There are two types of stages: [[ResultStage]], for the final stage that
 *    executes an action, and [[ShuffleMapStage]], which writes map output files for a shuffle.
 *    Stages are often shared across multiple jobs, if these jobs reuse the same RDDs.
 *
 *  - Tasks are individual units of work, each sent to one machine.
 *
 *  - Cache tracking: the DAGScheduler figures out which RDDs are cached to avoid recomputing them
 *    and likewise remembers which shuffle map stages have already produced output files to avoid
 *    redoing the map side of a shuffle.
 *
 *  - Preferred locations: the DAGScheduler also computes where to run each task in a stage based
 *    on the preferred locations of its underlying RDDs, or the location of cached or shuffle data.
 *
 *  - Cleanup: all data structures are cleared when the running jobs that depend on them finish,
 *    to prevent memory leaks in a long-running application.
 *
 * To recover from failures, the same stage might need to run multiple times, which are called
 * "attempts". If the TaskScheduler reports that a task failed because a map output file from a
 * previous stage was lost, the DAGScheduler resubmits that lost stage. This is detected through a
 * CompletionEvent with FetchFailed, or an ExecutorLost event. The DAGScheduler will wait a small
 * amount of time to see whether other nodes or tasks fail, then resubmit TaskSets for any lost
 * stage(s) that compute the missing tasks. As part of this process, we might also have to create
 * Stage objects for old (finished) stages where we previously cleaned up the Stage object. Since
 * tasks from the old attempt of a stage could still be running, care must be taken to map any
 * events received in the correct Stage object.
 *
 * Here's a checklist to use when making or reviewing changes to this class:
 *
 *  - All data structures should be cleared when the jobs involving them end to avoid indefinite
 *    accumulation of state in long-running programs.
 *
 *  - When adding a new data structure, update `DAGSchedulerSuite.assertDataStructuresEmpty` to
 *    include the new structure. This will help to catch memory leaks.
 */



/**
 * Submit an action job to the scheduler.
 *
 * @param rdd target RDD to run tasks on
 * @param func a function to run on each partition of the RDD
 * @param partitions set of partitions to run on; some jobs may not want to compute on all
 *   partitions of the target RDD, e.g. for operations like first()
 * @param callSite where in the user program this job was called
 * @param resultHandler callback to pass each result to
 * @param properties scheduler properties to attach to this job, e.g. fair scheduler pool name
 *
 * @return a JobWaiter object that can be used to block until the job finishes executing
 *         or can be used to cancel the job.
 *
 * @throws IllegalArgumentException when partitions ids are illegal
 */
def submitJob[T, U](
    rdd: RDD[T],
    func: (TaskContext, Iterator[T]) => U,
    partitions: Seq[Int],
    callSite: CallSite,
    resultHandler: (Int, U) => Unit,
    properties: Properties): JobWaiter[U] = {
  // Check to make sure we are not launching a task on a partition that does not exist.
  val maxPartitions = rdd.partitions.length
  partitions.find(p => p >= maxPartitions || p < 0).foreach { p =>
    throw new IllegalArgumentException(
      "Attempting to access a non-existent partition: " + p + ". " +
        "Total number of partitions: " + maxPartitions)
  }

  val jobId = nextJobId.getAndIncrement()
  if (partitions.size == 0) {
    // Return immediately if the job is running 0 tasks
    return new JobWaiter[U](this, jobId, 0, resultHandler)
  }

  assert(partitions.size > 0)
  val func2 = func.asInstanceOf[(TaskContext, Iterator[_]) => _]
  val waiter = new JobWaiter(this, jobId, partitions.size, resultHandler)
  eventProcessLoop.post(JobSubmitted(
    jobId, rdd, func2, partitions.toArray, callSite, waiter,
    SerializationUtils.clone(properties)))
  waiter
}

4.DAGSchedulerEventProcessLoop

eventProcessLoop是一个DAGSchedulerEventProcessLoop类对象，即一个DAG调度事件处理的监听。eventProcessLoop中调用doOnReceive来进行监听

本身这个类相当于一个调度事件处理的监听。实现如下：

private[scheduler] class DAGSchedulerEventProcessLoop(dagScheduler: DAGScheduler)
  extends EventLoop[DAGSchedulerEvent]("dag-scheduler-event-loop") with Logging {

  private[this] val timer = dagScheduler.metricsSource.messageProcessingTimer

  /**
   * The main event loop of the DAG scheduler.
   */
  override def onReceive(event: DAGSchedulerEvent): Unit = {
    val timerContext = timer.time()
    try {
      doOnReceive(event)
    } finally {
      timerContext.stop()
    }
  }

  private def doOnReceive(event: DAGSchedulerEvent): Unit = event match {
    case JobSubmitted(jobId, rdd, func, partitions, callSite, listener, properties) =>
      dagScheduler.handleJobSubmitted(jobId, rdd, func, partitions, callSite, listener, properties)

    case MapStageSubmitted(jobId, dependency, callSite, listener, properties) =>
      dagScheduler.handleMapStageSubmitted(jobId, dependency, callSite, listener, properties)

    case StageCancelled(stageId, reason) =>
      dagScheduler.handleStageCancellation(stageId, reason)

    case JobCancelled(jobId, reason) =>
      dagScheduler.handleJobCancellation(jobId, reason)

    case JobGroupCancelled(groupId) =>
      dagScheduler.handleJobGroupCancelled(groupId)

    case AllJobsCancelled =>
      dagScheduler.doCancelAllJobs()

    case ExecutorAdded(execId, host) =>
      dagScheduler.handleExecutorAdded(execId, host)

    case ExecutorLost(execId, reason) =>
      val filesLost = reason match {
        case SlaveLost(_, true) => true
        case _ => false
      }
      dagScheduler.handleExecutorLost(execId, filesLost)

    case BeginEvent(task, taskInfo) =>
      dagScheduler.handleBeginEvent(task, taskInfo)

    case GettingResultEvent(taskInfo) =>
      dagScheduler.handleGetTaskResult(taskInfo)

    case completion: CompletionEvent =>
      dagScheduler.handleTaskCompletion(completion)

    case TaskSetFailed(taskSet, reason, exception) =>
      dagScheduler.handleTaskSetFailed(taskSet, reason, exception)

    case ResubmitFailedStages =>
      dagScheduler.resubmitFailedStages()
  }

  override def onError(e: Throwable): Unit = {
    logError("DAGSchedulerEventProcessLoop failed; shutting down SparkContext", e)
    try {
      dagScheduler.doCancelAllJobs()
    } catch {
      case t: Throwable => logError("DAGScheduler failed to cancel all jobs.", t)
    }
    dagScheduler.sc.stopInNewThread()
  }

  override def onStop(): Unit = {
    // Cancel any active jobs in postStop hook
    dagScheduler.cleanUpAfterSchedulerStop()
  }
}

5.dagscheduler.handlejobsubmitted

最后我们来看一下handlejobsubmitted方法。至此job的提交就完成了。

private[scheduler] def handleJobSubmitted(jobId: Int,
    finalRDD: RDD[_],
    func: (TaskContext, Iterator[_]) => _,
    partitions: Array[Int],
    callSite: CallSite,
    listener: JobListener,
    properties: Properties) {
  var finalStage: ResultStage = null
  try {
    // New stage creation may throw an exception if, for example, jobs are run on a
    // HadoopRDD whose underlying HDFS files have been deleted.
    finalStage = createResultStage(finalRDD, func, partitions, jobId, callSite)
  } catch {
    case e: Exception =>
      logWarning("Creating new stage failed due to exception - job: " + jobId, e)
      listener.jobFailed(e)
      return
  }

  val job = new ActiveJob(jobId, finalStage, callSite, listener, properties)
  clearCacheLocs()
  logInfo("Got job %s (%s) with %d output partitions".format(
    job.jobId, callSite.shortForm, partitions.length))
  logInfo("Final stage: " + finalStage + " (" + finalStage.name + ")")
  logInfo("Parents of final stage: " + finalStage.parents)
  logInfo("Missing parents: " + getMissingParentStages(finalStage))

  val jobSubmissionTime = clock.getTimeMillis()
  jobIdToActiveJob(jobId) = job
  activeJobs += job
  finalStage.setActiveJob(job)
  val stageIds = jobIdToStageIds(jobId).toArray
  val stageInfos = stageIds.flatMap(id => stageIdToStage.get(id).map(_.latestInfo))
  listenerBus.post(
    SparkListenerJobStart(job.jobId, jobSubmissionTime, stageInfos, properties))
  submitStage(finalStage)
}

其中handleJobSubmitted定义了finalStage。我们下面来看看dagscheduler是如何划分stage的。

划分stage

正如之前提到的，stage的划分是根据宽窄依赖，来划分的。调用栈如下：

• DAGScheduler.createResultStage
  • DAGScheduler.getParentStagesAndId
    • DAGScheduler.getParentStages
      • DAGScheduler.getShuffleMapStage
        • DAGScheduler.getAncestorShuffleDependencies
        • DAGScheduler.newOrUsedShuffleStage
          • DAGScheduler.newShuffleMapStage

接下来，我们依次来分析：

1.createResultStage

Spark的Stage调用是从最后一个RDD所在的Stage，ResultStage开始划分的，这里即为G所在的Stage。但是在生成这个Stage之前会生成它的parent Stage，就这样递归的把parent Stage都先生成了。代码如下：

/**
 * Create a ResultStage associated with the provided jobId.
 */
private def createResultStage(
    rdd: RDD[_],
    func: (TaskContext, Iterator[_]) => _,
    partitions: Array[Int],
    jobId: Int,
    callSite: CallSite): ResultStage = {
  val parents = getOrCreateParentStages(rdd, jobId)
  val id = nextStageId.getAndIncrement()
  val stage = new ResultStage(id, rdd, func, partitions, parents, jobId, callSite)
  stageIdToStage(id) = stage
  updateJobIdStageIdMaps(jobId, stage)
  stage
}

2.getOrCreateParentStages

/**
 * Get or create the list of parent stages for a given RDD.  The new Stages will be created with
 * the provided firstJobId.
 */
private def getOrCreateParentStages(rdd: RDD[_], firstJobId: Int): List[Stage] = {
  getShuffleDependencies(rdd).map { shuffleDep =>
    getOrCreateShuffleMapStage(shuffleDep, firstJobId)
  }.toList
}

3.getShuffleDependencies

/**
 * Returns shuffle dependencies that are immediate parents of the given RDD.
 *
 * This function will not return more distant ancestors.  For example, if C has a shuffle
 * dependency on B which has a shuffle dependency on A:
 *
 * A <-- B <-- C
 *
 * calling this function with rdd C will only return the B <-- C dependency.
 *
 * This function is scheduler-visible for the purpose of unit testing.
 */
private[scheduler] def getShuffleDependencies(
    rdd: RDD[_]): HashSet[ShuffleDependency[_, _, _]] = {
  //存储parentstage
  val parents = new HashSet[ShuffleDependency[_, _, _]]
  //存储已经访问过的RDD
  val visited = new HashSet[RDD[_]]
  //存储需要被处理的rdd
  val waitingForVisit = new Stack[RDD[_]]
  waitingForVisit.push(rdd)
  while (waitingForVisit.nonEmpty) {
    val toVisit = waitingForVisit.pop()
    if (!visited(toVisit)) {
      //加入访问集合
      visited += toVisit
      //便利该rdd所有的依赖
      toVisit.dependencies.foreach {
        //若是款里来生成新的stage
        case shuffleDep: ShuffleDependency[_, _, _] =>
          parents += shuffleDep
        //窄依赖，加入stack，等待处理。
        case dependency =>
          waitingForVisit.push(dependency.rdd)
      }
    }
  }
  parents
}

4.getOrCreateShuffleMapStage

shuffleDep =>
  getOrCreateShuffleMapStage(shuffleDep, firstJobId)

上面提到，如果是宽依赖，就声称新的stage，其中parents += shuffleDep。shuffleDep是getOrCreateShuffleMapStage生成的

/**
 * Gets a shuffle map stage if one exists in shuffleIdToMapStage. Otherwise, if the
 * shuffle map stage doesn't already exist, this method will create the shuffle map stage in
 * addition to any missing ancestor shuffle map stages.
 */
private def getOrCreateShuffleMapStage(
    shuffleDep: ShuffleDependency[_, _, _],
    firstJobId: Int): ShuffleMapStage = {
  shuffleIdToMapStage.get(shuffleDep.shuffleId) match {
    case Some(stage) =>
      stage

    case None =>
      // Create stages for all missing ancestor shuffle dependencies.
     //检查这个stage的parent stage是否生成，没有生成他们 getMissingAncestorShuffleDependencies(shuffleDep.rdd).foreach { dep =>
        // Even though getMissingAncestorShuffleDependencies only returns shuffle dependencies
        // that were not already in shuffleIdToMapStage, it's possible that by the time we
        // get to a particular dependency in the foreach loop, it's been added to
        // shuffleIdToMapStage by the stage creation process for an earlier dependency. See
        // SPARK-13902 for more information.
        if (!shuffleIdToMapStage.contains(dep.shuffleId)) {
          createShuffleMapStage(dep, firstJobId)
        }
      }
      // Finally, create a stage for the given shuffle dependency.
      createShuffleMapStage(shuffleDep, firstJobId)
  }
}

接下来看到createShuffleMapStage

/**
 * Creates a ShuffleMapStage that generates the given shuffle dependency's partitions. If a
 * previously run stage generated the same shuffle data, this function will copy the output
 * locations that are still available from the previous shuffle to avoid unnecessarily
 * regenerating data.
 */
def createShuffleMapStage(shuffleDep: ShuffleDependency[_, _, _], jobId: Int): ShuffleMapStage = {
  val rdd = shuffleDep.rdd
  val numTasks = rdd.partitions.length
  val parents = getOrCreateParentStages(rdd, jobId)
  val id = nextStageId.getAndIncrement()
  val stage = new ShuffleMapStage(id, rdd, numTasks, parents, jobId, rdd.creationSite, shuffleDep)

  stageIdToStage(id) = stage
  shuffleIdToMapStage(shuffleDep.shuffleId) = stage
  updateJobIdStageIdMaps(jobId, stage)

  if (mapOutputTracker.containsShuffle(shuffleDep.shuffleId)) {
    // A previously run stage generated partitions for this shuffle, so for each output
    // that's still available, copy information about that output location to the new stage
    // (so we don't unnecessarily re-compute that data).
    val serLocs = mapOutputTracker.getSerializedMapOutputStatuses(shuffleDep.shuffleId)
    val locs = MapOutputTracker.deserializeMapStatuses(serLocs)
    (0 until locs.length).foreach { i =>
      if (locs(i) ne null) {
        // locs(i) will be null if missing
        stage.addOutputLoc(i, locs(i))
      }
    }
  } else {
    // Kind of ugly: need to register RDDs with the cache and map output tracker here
    // since we can't do it in the RDD constructor because # of partitions is unknown
    logInfo("Registering RDD " + rdd.id + " (" + rdd.getCreationSite + ")")
    mapOutputTracker.registerShuffle(shuffleDep.shuffleId, rdd.partitions.length)
  }
  stage
}

至此，stage划分结束。

以下内容版本貌似不太对，但整体思想是一样的。

• 首先，我们想 newResultStage RDD_G所在的Stage3
• 但在new Stage之前会调用getParentStagesAndId
• getParentStagesAndId中又会调用getParentStages，来广度优先的遍历RDD_G所依赖的RDD。如果是窄依赖，就纳入G所在的Stage3，如RDD_B就纳入了Stage3
• 若过是宽依赖，我们这里以RDD_F为例（与RDD_A处理过程相同）。我们就会调用getShuffleMapStage，来判断RDD_F所在的Stage2是否已经生成了，如果生成了就直接返回。
• 若还没生成，我们先调用getAncestorShuffleDependencies。getAncestorShuffleDependencies类似于getParentStages，也是用广度优先的遍历RDD_F所依赖的RDD。如果是窄依赖，如RDD_C、RDD_D和RDD_E，都被纳入了F所在的Stage2。但是假设RDD_E有个parent RDD ``RDD_H，RDD_H和RDD_E之间是宽依赖，那么该怎么办呢？我们会先判断RDD_H所在的Stage是否已经生成。若还没生成，我们把它put到一个parents Stack 中，最后返回。
• 对于那些返回的还没生成的Stage我们会调用newOrUsedShuffleStage
• newOrUsedShuffleStage会调用newShuffleMapStage，来生成新的Stage。而newShuffleMapStage的实现类似于newResultStage。这样我们就可以递归下去，使得每个Stage所依赖的Stage都已经生成了，再来生成这个的Stage。如这里，会将RDD_H所在的Stage生成了，然后在再生成Stage2。
• newOrUsedShuffleStage生成新的Stage后，会判断Stage是否被计算过。若已经被计算过，就从mapOutPutTracker中复制计算结果。若没计算过，则向mapOutPutTracker注册占位。
• 最后，回到newResultStage中，new ResultStage，这里即生成了Stage3。至此，Stage划分过程就结束了。

生成job

现在回过头看handleJobSubmitted方法，其中，创建finalStage后，我们要根据stage来创建job，

//生成job
val job = new ActiveJob(jobId, finalStage, callSite, listener, properties)
clearCacheLocs()
logInfo("Got job %s (%s) with %d output partitions".format(
  job.jobId, callSite.shortForm, partitions.length))
logInfo("Final stage: " + finalStage + " (" + finalStage.name + ")")
logInfo("Parents of final stage: " + finalStage.parents)
logInfo("Missing parents: " + getMissingParentStages(finalStage))
//获取job提交的时间
val jobSubmissionTime = clock.getTimeMillis()
//得到ID，添加到activeJobs
jobIdToActiveJob(jobId) = job
activeJobs += job
//吧finalstafe设置为该job
finalStage.setActiveJob(job)
//得到stage id
val stageIds = jobIdToStageIds(jobId).toArray
val stageInfos = stageIds.flatMap(id => stageIdToStage.get(id).map(_.latestInfo))
listenerBus.post(
  SparkListenerJobStart(job.jobId, jobSubmissionTime, stageInfos, properties))
//提交
submitStage(finalStage)

注意这里提交的stage，生成的job被设置为stage的activejob。接下来我们看一下stage是如何提交的。

提交stage

1.submitStage

正如注释写的那样，提交finalstage，我们需要看看哪些parentstage缺失。

/** Submits stage, but first recursively submits any missing parents. */
private def submitStage(stage: Stage) {
  val jobId = activeJobForStage(stage)
  if (jobId.isDefined) {
    logDebug("submitStage(" + stage + ")")
    if (!waitingStages(stage) && !runningStages(stage) && !failedStages(stage)) {
      //获取缺失的parent stage
      val missing = getMissingParentStages(stage).sortBy(_.id)
      logDebug("missing: " + missing)
      if (missing.isEmpty) {
        //如果没有缺失的parent stage，那么submitMissingTasks会完成最后一步，向taskscheduler提交task，
        logInfo("Submitting " + stage + " (" + stage.rdd + "), which has no missing parents")
        submitMissingTasks(stage, jobId.get)
      } else {
        //深度遍历。存在缺失stage，再来递归判断父亲stage是否缺失
        for (parent <- missing) {
          submitStage(parent)
        }
        waitingStages += stage
      }
    }
  } else {
    abortStage(stage, "No active job for stage " + stage.id, None)
  }
}

2.getMissingParentStages

getMissingParentStages类似于getParentStages，也是使用广度优先遍历

private def getMissingParentStages(stage: Stage): List[Stage] = {
  val missing = new HashSet[Stage]
  val visited = new HashSet[RDD[_]]
  // We are manually maintaining a stack here to prevent StackOverflowError
  // caused by recursively visiting
  val waitingForVisit = new Stack[RDD[_]]
  def visit(rdd: RDD[_]) {
    if (!visited(rdd)) {
      visited += rdd
      val rddHasUncachedPartitions = getCacheLocs(rdd).contains(Nil)
      if (rddHasUncachedPartitions) {
        for (dep <- rdd.dependencies) {
          dep match {
            case shufDep: ShuffleDependency[_, _, _] =>
              val mapStage = getOrCreateShuffleMapStage(shufDep, stage.firstJobId)
              if (!mapStage.isAvailable) {
                missing += mapStage
              }
            case narrowDep: NarrowDependency[_] =>
              waitingForVisit.push(narrowDep.rdd)
          }
        }
      }
    }
  }
  waitingForVisit.push(stage.rdd)
  while (waitingForVisit.nonEmpty) {
    visit(waitingForVisit.pop())
  }
  missing.toList
}

3.submitMissingTasks

最后来看一下，submitMissingTasks如何完成最后的工作，提交task。

/** Called when stage's parents are available and we can now do its task. */
private def submitMissingTasks(stage: Stage, jobId: Int) {
  logDebug("submitMissingTasks(" + stage + ")")

  // First figure out the indexes of partition ids to compute.
  val partitionsToCompute: Seq[Int] = stage.findMissingPartitions()

  // Use the scheduling pool, job group, description, etc. from an ActiveJob associated
  // with this Stage
  val properties = jobIdToActiveJob(jobId).properties

  runningStages += stage
  // SparkListenerStageSubmitted should be posted before testing whether tasks are
  // serializable. If tasks are not serializable, a SparkListenerStageCompleted event
  // will be posted, which should always come after a corresponding SparkListenerStageSubmitted
  // event.
  stage match {
    case s: ShuffleMapStage =>
      outputCommitCoordinator.stageStart(stage = s.id, maxPartitionId = s.numPartitions - 1)
    case s: ResultStage =>
      outputCommitCoordinator.stageStart(
        stage = s.id, maxPartitionId = s.rdd.partitions.length - 1)
  }
  val taskIdToLocations: Map[Int, Seq[TaskLocation]] = try {
    stage match {
      case s: ShuffleMapStage =>
        partitionsToCompute.map { id => (id, getPreferredLocs(stage.rdd, id))}.toMap
      case s: ResultStage =>
        partitionsToCompute.map { id =>
          val p = s.partitions(id)
          (id, getPreferredLocs(stage.rdd, p))
        }.toMap
    }
  } catch {
    case NonFatal(e) =>
      stage.makeNewStageAttempt(partitionsToCompute.size)
      listenerBus.post(SparkListenerStageSubmitted(stage.latestInfo, properties))
      abortStage(stage, s"Task creation failed: $e\n${Utils.exceptionString(e)}", Some(e))
      runningStages -= stage
      return
  }

  stage.makeNewStageAttempt(partitionsToCompute.size, taskIdToLocations.values.toSeq)
  listenerBus.post(SparkListenerStageSubmitted(stage.latestInfo, properties))

  // TODO: Maybe we can keep the taskBinary in Stage to avoid serializing it multiple times.
  // Broadcasted binary for the task, used to dispatch tasks to executors. Note that we broadcast
  // the serialized copy of the RDD and for each task we will deserialize it, which means each
  // task gets a different copy of the RDD. This provides stronger isolation between tasks that
  // might modify state of objects referenced in their closures. This is necessary in Hadoop
  // where the JobConf/Configuration object is not thread-safe.
  var taskBinary: Broadcast[Array[Byte]] = null
  try {
    // For ShuffleMapTask, serialize and broadcast (rdd, shuffleDep).
    // For ResultTask, serialize and broadcast (rdd, func).
    //对于最后一个任务resulttask，序列化并广播
    val taskBinaryBytes: Array[Byte] = stage match {
      case stage: ShuffleMapStage =>
        JavaUtils.bufferToArray(
          closureSerializer.serialize((stage.rdd, stage.shuffleDep): AnyRef))
      case stage: ResultStage =>
        JavaUtils.bufferToArray(closureSerializer.serialize((stage.rdd, stage.func): AnyRef))
    }

    taskBinary = sc.broadcast(taskBinaryBytes)
  } catch {
    // In the case of a failure during serialization, abort the stage.
    case e: NotSerializableException =>
      abortStage(stage, "Task not serializable: " + e.toString, Some(e))
      runningStages -= stage

      // Abort execution
      return
    case NonFatal(e) =>
      abortStage(stage, s"Task serialization failed: $e\n${Utils.exceptionString(e)}", Some(e))
      runningStages -= stage
      return
  }

  val tasks: Seq[Task[_]] = try {
    val serializedTaskMetrics = closureSerializer.serialize(stage.latestInfo.taskMetrics).array()
    stage match {
      //如果是其他的stage，创建ShuffleMapTask
      case stage: ShuffleMapStage =>
        stage.pendingPartitions.clear()
        partitionsToCompute.map { id =>
          val locs = taskIdToLocations(id)
          val part = stage.rdd.partitions(id)
          stage.pendingPartitions += id
          new ShuffleMapTask(stage.id, stage.latestInfo.attemptId,
            taskBinary, part, locs, properties, serializedTaskMetrics, Option(jobId),
            Option(sc.applicationId), sc.applicationAttemptId)
        }
    //最后的stage 创建ResultTask
      case stage: ResultStage =>
        partitionsToCompute.map { id =>
          val p: Int = stage.partitions(id)
          val part = stage.rdd.partitions(p)
          val locs = taskIdToLocations(id)
          new ResultTask(stage.id, stage.latestInfo.attemptId,
            taskBinary, part, locs, id, properties, serializedTaskMetrics,
            Option(jobId), Option(sc.applicationId), sc.applicationAttemptId)
        }
    }
  } catch {
    case NonFatal(e) =>
      abortStage(stage, s"Task creation failed: $e\n${Utils.exceptionString(e)}", Some(e))
      runningStages -= stage
      return
  }

  if (tasks.size > 0) {
    logInfo(s"Submitting ${tasks.size} missing tasks from $stage (${stage.rdd}) (first 15 " +
      s"tasks are for partitions ${tasks.take(15).map(_.partitionId)})")
    //创建taskset，并提交，其中taskset的大侠根据tasks的数量决定。tasks的数量根据partiton的个数和数据决定
    taskScheduler.submitTasks(new TaskSet(
      tasks.toArray, stage.id, stage.latestInfo.attemptId, jobId, properties))
    stage.latestInfo.submissionTime = Some(clock.getTimeMillis())
  } else {
    // Because we posted SparkListenerStageSubmitted earlier, we should mark
    // the stage as completed here in case there are no tasks to run
    markStageAsFinished(stage, None)

    val debugString = stage match {
      case stage: ShuffleMapStage =>
        s"Stage ${stage} is actually done; " +
          s"(available: ${stage.isAvailable}," +
          s"available outputs: ${stage.numAvailableOutputs}," +
          s"partitions: ${stage.numPartitions})"
      case stage : ResultStage =>
        s"Stage ${stage} is actually done; (partitions: ${stage.numPartitions})"
    }
    logDebug(debugString)

    submitWaitingChildStages(stage)
  }
}

参考：http://blog.csdn.net/u011239443/article/details/53911902

总结： 总的来说dagscheduler是来划分dag变成多个stage。每一个stage，生成一个taskset。每一个taskset包含一些列task，每个tash作用于每一个partition。其中讨论了 如何生成一个job？ 如何提交job？ job中如何划分stage？ 如何提交stage？ 如何提交taskset到taskscheduler？