Spark算子执行流程详解之二
来源:互联网 发布:鸡兔同笼用c语言编程 编辑:程序博客网 时间:2024/05/29 14:54
4.count
def count(): Long = sc.runJob(this, Utils.getIteratorSize_).sum
计算数据总量,每个分区各自计算自己的总数,然后汇总到driver端,driver端再把每个分区的总数相加统计出对应rdd的数据量,其流程如下:
5.countApprox
在一定的超时时间之内返回rdd元素的个数,其rdd元素的总数分布符合正态分布,其分布因子为confidence,当超过timeout时,返回一个未完成的结果。
/**
* :: Experimental ::
* Approximate version of count() that returns a potentially incomplete result
* within a timeout, even if not all tasks have finished.
*/
@Experimental
def countApprox(
timeout: Long,
confidence: Double = 0.95): PartialResult[BoundedDouble] = withScope {
//定义在excutor端计算总数的函数
val countElements: (TaskContext, Iterator[T]) => Long = { (ctx, iter) =>
var result = 0L
while (iter.hasNext) {
result += 1L
iter.next()
}
result
}
//定义在driver端的一个监听回调函数,当task完成的时候,会触发里面的merge操作,当超时时间到之后或者任务提前完成的话,会取里面的当前状态,即currentResult
val evaluator = newCountEvaluator(partitions.length, confidence)
//提交任务
sc.runApproximateJob(this, countElements, evaluator, timeout)
}
继续往下看,看看evaluator是如何执行的:
def runApproximateJob[T,U,R](
rdd: RDD[T],
func: (TaskContext, Iterator[T]) =>U,
evaluator: ApproximateEvaluator[U, R],
timeout: Long): PartialResult[R] = {
assertNotStopped()
val callSite = getCallSite
logInfo("Starting job: " + callSite.shortForm)
val start = System.nanoTime
val cleanedFunc = clean(func)
// cleanedFunc就是countElements,evaluator就是CountEvaluator,超时时间为timeout
val result = dagScheduler.runApproximateJob(rdd, cleanedFunc, evaluator, callSite, timeout,
localProperties.get)
logInfo(
"Job finished: " + callSite.shortForm +", took "+ (System.nanoTime- start) / 1e9+ " s")
result
}
继续看runApproximateJob的实现:
def runApproximateJob[T,U,R](
rdd: RDD[T],
func: (TaskContext, Iterator[T]) =>U,
evaluator: ApproximateEvaluator[U, R],
callSite: CallSite,
timeout: Long,
properties: Properties): PartialResult[R] = {
//定义一个监听器,当有任务完成的时候触发taskSucceeded,当超时时间到的时候返回CountEvaluator的当前值
val listener = newApproximateActionListener(rdd, func, evaluator, timeout)
val func2 = func.asInstanceOf[(TaskContext,Iterator[_]) => _]
val partitions = (0until rdd.partitions.size).toArray
val jobId = nextJobId.getAndIncrement()
//提交任务
eventProcessLoop.post(JobSubmitted(
jobId, rdd, func2, partitions, allowLocal = false, callSite, listener,
SerializationUtils.clone(properties)))
//等待计算结果
listener.awaitResult() // Will throw an exception if the job fails
}
因此其超时计算总数的逻辑主要在ApproximateActionListener里面,请看ApproximateActionListener:
private[spark] classApproximateActionListener[T,U, R](
rdd: RDD[T],
func: (TaskContext, Iterator[T]) =>U,
evaluator: ApproximateEvaluator[U, R],
timeout: Long)
extends JobListener {
val startTime= System.currentTimeMillis()
val totalTasks= rdd.partitions.size
var finishedTasks=0
var failure: Option[Exception] = None // Set if the job has failed (permanently)
var resultObject: Option[PartialResult[R]] = None// Set if we've already returned a PartialResult
//当某个分区完成的时候触发taskSucceeded回调函数
override def taskSucceeded(index: Int, result: Any) {
synchronized {
//更新CountEvaluator的当前值
evaluator.merge(index, result.asInstanceOf[U])
finishedTasks += 1
if (finishedTasks==totalTasks) {//当全部分区都完成的是退出等待,返回计算结果
// If we had already returned a PartialResult, set its final value
resultObject.foreach(r => r.setFinalValue(evaluator.currentResult()))
// Notify any waiting thread that may have called awaitResult
//退出等待
this.notifyAll()
}
}
}
……
/**
* Waits for up to timeout milliseconds since the listener was created and then returns a
* PartialResult with the result so far. This may be complete if the whole job is done.
*/
//等待计算结果
def awaitResult(): PartialResult[R] = synchronized {
val finishTime = startTime+ timeout
while (true) {
val time = System.currentTimeMillis()
if (failure.isDefined) {
throw failure.get
} else if (finishedTasks==totalTasks) {//如果在超时时间之内计算完成,则返回全部结果
return new PartialResult(evaluator.currentResult(),true)
} else if (time >= finishTime) {//如果已经超时,则返回部分结果
resultObject = Some(newPartialResult(evaluator.currentResult(),false))
return resultObject.get
} else {//如果超时时间没到,则继续休眠
this.wait(finishTime - time)
}
}
// Should never be reached, but required to keep the compiler happy
return null
}
}
其中如果在超时时间之内没有完成的话,evaluator.currentResult()会返回符合总数符合正态分布的一个近似结果,感兴趣的同学可以继续研究下去:
private[spark] classCountEvaluator(totalOutputs: Int, confidence: Double)
extends ApproximateEvaluator[Long, BoundedDouble] {
var outputsMerged=0
var sum: Long =0
override def merge(outputId: Int, taskResult: Long) {
outputsMerged += 1
sum += taskResult
}
override def currentResult(): BoundedDouble = {
if (outputsMerged== totalOutputs) {//全部完成
new BoundedDouble(sum,1.0,sum,sum)
} else if (outputsMerged==0) {//一个任务都没完成
new BoundedDouble(0,0.0, Double.NegativeInfinity, Double.PositiveInfinity)
} else {//部分完成,计算其理论总数的正态分布参数
val p = outputsMerged.toDouble / totalOutputs
val mean = (sum+1 - p) / p
val variance = (sum+1) * (1 - p) / (p * p)
val stdev = math.sqrt(variance)
val confFactor = newNormalDistribution().
inverseCumulativeProbability(1 - (1- confidence) /2)
val low = mean - confFactor * stdev
val high = mean + confFactor * stdev
new BoundedDouble(mean, confidence, low, high)
}
}
}
因此countApprox的计算过程大致如下:1)excutor端不断的计算分区的总数然后上报给driver端;2)driver端接受excutor上报的总数进行统计,如果在超时时间之内没有全部上报完成的话,则强制退出,返回一个其总数符合正态分布的值,如果在超时时间之内计算完成的话,则返回一个准确值。
6.countApproxDistinct
作用是对RDD集合内容进行去重统计,该统计是一个大约的统计,参数relativeSD控制统计的精确度。relativeSD越小,结果越准确。
/**
* Return approximate number of distinct elements in the RDD.
*
* The algorithm used is based on streamlib's implementation of "HyperLogLog in Practice:
* Algorithmic Engineering of a State of The Art Cardinality Estimation Algorithm", available
* <a href="http://dx.doi.org/10.1145/2452376.2452456">here</a>.
*
* @param relativeSD Relative accuracy. Smaller values create counters that require more space.
* It must be greater than 0.000017.
*/
def countApproxDistinct(relativeSD: Double =0.05): Long = withScope {
require(relativeSD > 0.000017, s"accuracy ($relativeSD) must be greater than 0.000017")
val p = math.ceil(2.0* math.log(1.054/ relativeSD) / math.log(2)).toInt
countApproxDistinct(if (p < 4) 4 elsep,0)
}
采用的是HyperLogLog in Practice算法,原理比较深奥,有兴趣的可以深究。
实例如下:
object CountApproxDistinct {
def main(args: Array[String]) {
val conf = new SparkConf().setAppName("spark-demo").setMaster("local")
val sc = new SparkContext(conf)
/**
* 构建一个集合,分成20个partition
*/
val a = sc.parallelize(1 to 10000 , 20)
//RDD a内容复制5遍,其中有50000个元素
val b = a++a++a++a++a
//结果是9760,不传参数,默认是0.05
println(b.countApproxDistinct())
//结果是9760
println(b.countApproxDistinct(0.05))
//8224
println(b.countApproxDistinct(0.1))
//10000
println(b.countApproxDistinct(0.001))
}
}
7.collect
def collect(): Array[T] = withScope {
val results = sc.runJob(this, (iter:Iterator[T]) => iter.toArray)
Array.concat(results: _*)
}
获取Rdd的所有数据,然后缓存在Driver端,其流程如下:
如果RDD数据量很大的话,请谨慎使用,因为会缓存该RDD的所有数据量。
8.toLocalIterator
返回一个保护所有记录的迭代器
/**
* Return an iterator that contains all of the elements in this RDD.
*
* The iterator will consume as much memory as the largest partition in this RDD.
*
* Note: this results in multiple Spark jobs, and if the input RDD is the result
* of a wide transformation (e.g. join with different partitioners), to avoid
* recomputing the input RDD should be cached first.
*/
def toLocalIterator:Iterator[T] = withScope {
//针对每个分区触发一次action
def collectPartition(p: Int): Array[T] = {
sc.runJob(this, (iter: Iterator[T]) => iter.toArray, Seq(p), allowLocal =false).head
}
//调用flatMap将所有记录组装起来返回单个迭代器
(0 until partitions.length).iterator.flatMap(i => collectPartition(i))
}
即:
scala> val rdd = sc.parallelize(1 to 10,2)
rdd: org.apache.spark.rdd.RDD[Int] = ParallelCollectionRDD[0] at parallelize at <console>:24
scala> val it = rdd.toLocalIterator
it: Iterator[Int] = non-empty iterator
scala> while(it.hasNext){
| println(it.next)
| }
1
2
3
4
5
6
7
8
9
10
9.takeOrdered
takeOrdered函数用于从RDD中,按照默认(升序)或指定排序规则,返回前num个元素。
def takeOrdered(num: Int)(implicitord:Ordering[T]): Array[T] = withScope {
if (num == 0) {
Array.empty
} else {
val mapRDDs = mapPartitions { items =>
//先在excutor端进行排序,按照ord排序规则,转化为前num个优先队列
// Priority keeps the largest elements, so let's reverse the ordering.
val queue = new BoundedPriorityQueue[T](num)(ord.reverse)
queue ++= util.collection.Utils.takeOrdered(items, num)(ord)
Iterator.single(queue)
}
if (mapRDDs.partitions.length ==0) {
Array.empty
} else {
//将分区的计算结果传送给driver,转化为数组,进行排序取前num条记录
mapRDDs.reduce { (queue1, queue2) =>
queue1 ++= queue2
queue1
}.toArray.sorted(ord)
}
}
}
例如:
List<Integer> data = Arrays.asList(1,4,3,2,5,6);
JavaRDD<Integer> JavaRDD = jsc.parallelize(data,2);
for(Integer integer:JavaRDD.takeOrdered(2)){
System.out.println(integer);
}
打印
1
2
其执行流程如下:
- Spark算子执行流程详解之二
- Spark算子执行流程详解之三
- Spark算子执行流程详解之四
- Spark算子执行流程详解之五
- Spark算子执行流程详解之六
- Spark算子执行流程详解之七
- Spark算子执行流程详解之八
- Spark算子执行流程详解之一
- Spark算子(二)
- spark 算子combineByKey 详解
- spark 常用算子 详解
- Spark转换算子和执行算子
- spark笔记二之Spark程序模型Transformation算子与action算子
- Spark 之 算子
- spark之action算子
- spark算子之map_mapPartitions_mapPartitionsWithIndex
- spark算子实战(二)
- spark中算子详解:aggregateByKey
- eblearn 框架的理解学习(1)
- Quartz中时间表达式详细介绍与样例分析
- ASCII码表_全_完整版
- Android动画效果translate、scale、alpha、rotate详解
- 【ORACLE】ORA-01034ORACLE not available
- Spark算子执行流程详解之二
- 提高C#编程水平的50个要点 之十四“用多个构造函数时,利用构造函数链”
- Unity协程(Coroutine)原理剖析(一)
- 数学——Div算子
- <meta name="viewport" content="width=device-width, initial-scale=1.0, user-scalable=no, minimum-scal
- Windows远程桌面连接失败和无法复制粘贴
- Vue自定义指令
- Eclipse 调试笔记
- ubuntu 12.04 设置远程共享桌面