spark学习四 RDD转换是什么以及任务的运行

准备

1.   spark已经安装完毕

2.   spark运行在local modelocal-cluster mode

local-cluster mode

local-cluster模式也称为伪分布式,可以使用如下指令运行

MASTER=local[1,2,1024] bin/spark-shell
 

[1,2,1024] 分别表示,executor number, core number和内存大小,其中内存大小不应小于默认的512M

Driver Programme的初始化过程分析

初始化过程的涉及的主要源文件

1.   SparkContext.scala      整个初始化过程的入口

2.   SparkEnv.scala      创建BlockManager,MapOutputTrackerMaster, ConnectionManager, CacheManager

3.   DAGScheduler.scala      任务提交的入口,即将Job划分成各个stage的关键

4.   TaskSchedulerImpl.scala决定每个stage可以运行几个task,每个task分别在哪个executor上运行

5.   SchedulerBackend

1.   最简单的单机运行模式的话,看LocalBackend.scala

2.   如果是集群模式,看源文件SparkDeploySchedulerBackend

初始化过程步骤详解

步骤1根据初始化入参生成SparkConf,再根据SparkConf来创建SparkEnv, SparkEnv中主要包含以下关键性组件 1. BlockManager 2. MapOutputTracker 3. ShuffleFetcher 4.ConnectionManager

 private[spark] val env = SparkEnv.create(conf,  "",
    conf.get("spark.driver.host"),
    conf.get("spark.driver.port").toInt,
    isDriver = true, 
    isLocal = isLocal)
  SparkEnv.set(env)


 

步骤2:创建TaskScheduler,根据Spark的运行模式来选择相应的SchedulerBackend,同时启动taskscheduler,这一步至为关键

  private[spark] var taskScheduler = SparkContext.createTaskScheduler(this, master, appName)
  taskScheduler.start()


 

TaskScheduler.start目的是启动相应的SchedulerBackend,并启动定时器进行检测

overridedef start() {
    backend.start()
 
    if (!isLocal && conf.getBoolean("spark.speculation", false)) {
      logInfo("Starting speculative execution thread")
      import sc.env.actorSystem.dispatcher
      sc.env.actorSystem.scheduler.schedule(SPECULATION_INTERVAL milliseconds,
            SPECULATION_INTERVAL milliseconds) {
        checkSpeculatableTasks()
      }
    }
  }


 

步骤3:以上一步中创建的TaskScheduler实例为入参创建DAGScheduler并启动运行

@volatileprivate[spark] var dagScheduler = new DAGScheduler(taskScheduler)
  dagScheduler.start()
 


步骤4:启动WEB UI

ui.start()


 

RDD的转换过程

还是以最简单的wordcount为例说明rdd的转换过程

sc.textFile("README.md").flatMap(line=>line.split(" ")).map(word => (word, 1)).reduceByKey(_ + _)


上述一行简短的代码其实发生了很复杂的RDD转换,下面仔细解释每一步的转换过程和转换结果

步骤1:val rawFile = sc.textFile("README.md")

textFile先是生成hadoopRDD,然后再通过map操作生成MappedRDD,如果在spark-shell中执行上述语句,得到的结果可以证明所做的分析

scala> sc.textFile("README.md")
14/04/2313:11:48 WARN SizeEstimator: Failed to check whether UseCompressedOops is set; assuming yes
14/04/2313:11:48 INFO MemoryStore: ensureFreeSpace(119741) called with curMem=0, maxMem=311387750
14/04/2313:11:48 INFO MemoryStore: Block broadcast_0 stored as values to memory (estimated size 116.9 KB, free 296.8 MB)
14/04/2313:11:48 DEBUG BlockManager: Put block broadcast_0 locally took  277 ms
14/04/2313:11:48 DEBUG BlockManager: Put for block broadcast_0 without replication took  281 ms
res0: org.apache.spark.rdd.RDD[String] = MappedRDD[1] at textFile at :13

 

步骤2: val splittedText = rawFile.flatMap(line => line.split(""))

flatMap将原来的MappedRDD转换成为FlatMappedRDD

 def flatMap[U: ClassTag](f: T => TraversableOnce[U]): RDD[U] =   new FlatMappedRDD(this, sc.clean(f))
 

步骤3:val wordCount = splittedText.map(word => (word, 1))

利用word生成相应的键值对,上一步的FlatMappedRDD被转换成为MappedRDD

步骤4:val reduceJob = wordCount.reduceByKey(_ + _),这一步最复杂

步骤23中使用到的operation全部定义在RDD.scala中,而这里使用到的reduceByKey却在RDD.scala中见不到踪迹。reduceByKey的定义出现在源文件PairRDDFunctions.scala

细心的你一定会问reduceByKey不是MappedRDD的属性和方法啊,怎么能被MappedRDD调用呢?其实这背后发生了一个隐式的转换,该转换将MappedRDD转换成为PairRDDFunctions

implicit def rddToPairRDDFunctions[K: ClassTag, V: ClassTag](rdd: RDD[(K, V)]) = new PairRDDFunctions(rdd)


 

这种隐式的转换是scala的一个语法特征,如果想知道的更多,请用关键字"scalaimplicit method"进行查询,会有不少的文章对此进行详尽的介绍。

接下来再看一看reduceByKey的定义

  def reduceByKey(func: (V, V) => V): RDD[(K, V)] = {
    reduceByKey(defaultPartitioner(self), func)
  }
  def reduceByKey(partitioner: Partitioner, func: (V, V) => V): RDD[(K, V)] = {
    combineByKey[V]((v: V) => v, func, func, partitioner)
  }
 
  def combineByKey[C](createCombiner: V => C,
      mergeValue: (C, V) => C,
      mergeCombiners: (C, C) => C,
      partitioner: Partitioner,
      mapSideCombine: Boolean = true,
      serializerClass: String = null): RDD[(K, C)] = {
    if (getKeyClass().isArray) {
      if (mapSideCombine) {
        thrownew SparkException("Cannot use map-side combining with array keys.")
      }
      if (partitioner.isInstanceOf[HashPartitioner]) {
        thrownew SparkException("Default partitioner cannot partition array keys.")
      }
    }
    val aggregator = new Aggregator[K, V, C](createCombiner, mergeValue, mergeCombiners)
    if (self.partitioner == Some(partitioner)) {
      self.mapPartitionsWithContext((context, iter) => {
        new InterruptibleIterator(context, aggregator.combineValuesByKey(iter, context))
      }, preservesPartitioning = true)
    } elseif (mapSideCombine) {
      val combined = self.mapPartitionsWithContext((context, iter) => {
        aggregator.combineValuesByKey(iter, context)
      }, preservesPartitioning = true)
      val partitioned = new ShuffledRDD[K, C, (K, C)](combined, partitioner)
        .setSerializer(serializerClass)
      partitioned.mapPartitionsWithContext((context, iter) => {
        new InterruptibleIterator(context, aggregator.combineCombinersByKey(iter, context))
      }, preservesPartitioning = true)
    } else {
      // Don't apply map-side combiner.
      val values = new ShuffledRDD[K, V, (K, V)](self, partitioner).setSerializer(serializerClass)
      values.mapPartitionsWithContext((context, iter) => {
        new InterruptibleIterator(context, aggregator.combineValuesByKey(iter, context))
      }, preservesPartitioning = true)
    }
  }


 

reduceByKey最终会调用combineByKey,在这个函数中PairedRDDFunctions会被转换成为ShuffleRDD,当调用mapPartitionsWithContext之后,shuffleRDD被转换成为MapPartitionsRDD

Log输出能证明我们的分析

res1: org.apache.spark.rdd.RDD[(String, Int)] = MapPartitionsRDD[8] at reduceByKey at :13


 

RDD转换小结

小结一下整个RDD转换过程

HadoopRDD->MappedRDD->FlatMappedRDD->MappedRDD->PairRDDFunctions->ShuffleRDD->MapPartitionsRDD

整个转换过程好长啊,这一切的转换都发生在任务提交之前。

所有的RDD操作都是在做逻辑上的计算,只有在通过下面讲的DAGScheduler的处理生成对应的stage并划分出具体的taskset的时候才会触发真正的的数据操作

运行过程分析

数据集操作分类

在对任务运行过程中的函数调用关系进行分析之前,我们也来探讨一个偏理论的东西,作用于RDD之上的Transformantion为什么会是这个样子?

对这个问题的解答和数学搭上关系了,从理论抽象的角度来说,任务处理都可归结为“input->processing->output"inputoutput对应于数据集dataset.

在此基础上作一下简单的分类

1.   one-one一个dataset在转换之后还是一个dataset,而且datasetsize不变,如map

2.   one-one一个dataset在转换之后还是一个dataset,但size发生更改,这种更改有两种可能:扩大或缩小,如flatMapsize增大的操作,而subtractsize变小的操作

3.   many-one多个dataset合并为一个dataset,如combine, join

4.   one-many一个dataset分裂为多个dataset,groupBy

Task运行期的函数调用

task的提交过程参考本系列中的第二篇文章。本节主要讲解当task在运行期间是如何一步步调用到作用于RDD上的各个operation

·        TaskRunner.run

·        Task.run

·        Task.runTask (Task是一个基类,有两个子类,分别为ShuffleMapTaskResultTask)

·        RDD.iterator

·        RDD.computeOrReadCheckpoint

·        RDD.compute

或许当看到RDD.compute函数定义时,还是觉着f没有被调用,MappedRDDcompute定义为例

  overridedef compute(split: Partition, context: TaskContext) =  firstParent[T].iterator(split, context).map(f)  
 


注意,这里最容易产生错觉的地方就是map函数,这里的map不是RDD中的map而是scala中定义的iterator的成员函数map,请自行参考http://www.scala-lang.org/api/2.10.4/index.html#scala.collection.Iterator

堆栈输出

80         at org.apache.spark.rdd.HadoopRDD.getJobConf(HadoopRDD.scala:111)
81         at org.apache.spark.rdd.HadoopRDD$$anon$1.(HadoopRDD.scala:154)
82         at org.apache.spark.rdd.HadoopRDD.compute(HadoopRDD.scala:149)
83         at org.apache.spark.rdd.HadoopRDD.compute(HadoopRDD.scala:64)
84         at org.apache.spark.rdd.RDD.computeOrReadCheckpoint(RDD.scala:241)
85         at org.apache.spark.rdd.RDD.iterator(RDD.scala:232)
86         at org.apache.spark.rdd.MappedRDD.compute(MappedRDD.scala:31)
87         at org.apache.spark.rdd.RDD.computeOrReadCheckpoint(RDD.scala:241)
88         at org.apache.spark.rdd.RDD.iterator(RDD.scala:232)
89         at org.apache.spark.rdd.FlatMappedRDD.compute(FlatMappedRDD.scala:33)
90         at org.apache.spark.rdd.RDD.computeOrReadCheckpoint(RDD.scala:241)
91         at org.apache.spark.rdd.RDD.iterator(RDD.scala:232)
92         at org.apache.spark.rdd.MappedRDD.compute(MappedRDD.scala:31)
93         at org.apache.spark.rdd.RDD.computeOrReadCheckpoint(RDD.scala:241)
94         at org.apache.spark.rdd.RDD.iterator(RDD.scala:232)
95         at org.apache.spark.rdd.MapPartitionsRDD.compute(MapPartitionsRDD.scala:34)
96         at org.apache.spark.rdd.RDD.computeOrReadCheckpoint(RDD.scala:241)
97         at org.apache.spark.rdd.RDD.iterator(RDD.scala:232)
98         at org.apache.spark.scheduler.ShuffleMapTask.runTask(ShuffleMapTask.scala:161)
99         at org.apache.spark.scheduler.ShuffleMapTask.runTask(ShuffleMapTask.scala:102)
100         at org.apache.spark.scheduler.Task.run(Task.scala:53)
101         at org.apache.spark.executor.Executor$TaskRunner$$anonfun$run$1.apply$mcV$sp(Executor.scala:211)
 


ResultTask

compute的计算过程对于ShuffleMapTask比较复杂,绕的圈圈比较多,对于ResultTask就直接许多。

overridedef runTask(context: TaskContext): U = {
    metrics = Some(context.taskMetrics)
    try {
      func(context, rdd.iterator(split, context))
    } finally {
      context.executeOnCompleteCallbacks()
    }
  } 
 


 计算结果的传递

上面的分析知道,wordcount这个jo交之后,被DAGScheduler分为两个stage,第一个Stage是shuffleMapTask,第二个Stage是ResultTask.

那么ShuffleMapTask的计算结果是如何被ResultTask取得的呢?这个过程简述如下

1.   ShffuleMapTask将计算的状态(注意不是具体的数据)包装为MapStatus返回给DAGScheduler(不可能包装真正的数据的

2.   DAGSchedulerMapStatus保存到MapOutputTrackerMaster

3.   ResultTask在执行到ShuffleRDD时会调用BlockStoreShuffleFetcherfetch方法去获取数据(具体操作):

1.   第一件事就是咨询MapOutputTrackerMaster所要取的数据的location     验证上述2

2.   根据返回的结果调用BlockManager.getMultiple获取真正的数据

BlockStoreShuffleFetcherfetch函数伪码

    val blockManager = SparkEnv.get.blockManager
 
    val startTime = System.currentTimeMillis
    val statuses = SparkEnv.get.mapOutputTracker.getServerStatuses(shuffleId, reduceId)
    logDebug("Fetching map output location for shuffle %d, reduce %d took %d ms".format(
      shuffleId, reduceId, System.currentTimeMillis - startTime))
 
    val blockFetcherItr = blockManager.getMultiple(blocksByAddress, serializer)
    val itr = blockFetcherItr.flatMap(unpackBlock) 
 


注意上述代码中的getServerStatusesgetMultiple,一个是询问数据的位置,一个是去获取真正的数据。

 

你可能感兴趣的:(spark)