在 RDD 上使用 take 方法时，Apache Spark 引发反序列化错误

Question

我是 Spark 的新手，我正在使用 Scala 2.12.8 和 Spark 2.4.0。我正在尝试使用 Spark MLLib 中的随机森林分类器。我可以构建和训练分类器，分类器可以预测我是否在生成的 RDD 上使用 first() 函数。但是，如果我尝试使用 take(n) 函数，我会得到一个非常大、丑陋的堆栈跟踪。有谁知道我做错了什么？错误发生在以下行：“.take(3)”。我知道这是我在 RDD 上执行的第一个有效操作，所以如果有人能向我解释为什么它失败以及如何修复它，我将非常感激。

object ItsABreeze {
  def main(args: Array[String]): Unit = {
    val spark: SparkSession = SparkSession
      .builder()
      .appName("test")
      .getOrCreate()

    //Do stuff to file
    val data: RDD[LabeledPoint] = MLUtils.loadLibSVMFile(spark.sparkContext, "file.svm")

    // Split the data into training and test sets (30% held out for testing)
    val splits: Array[RDD[LabeledPoint]] = data.randomSplit(Array(0.7, 0.3))
    val (trainingData, testData) = (splits(0), splits(1))

    // Train a RandomForest model.
    // Empty categoricalFeaturesInfo indicates all features are continuous
    val numClasses = 4
    val categoricaFeaturesInfo = Map[Int, Int]()
    val numTrees = 3
    val featureSubsetStrategy = "auto"
    val impurity = "gini"
    val maxDepth = 5
    val maxBins = 32

    val model: RandomForestModel = RandomForest.trainClassifier(
      trainingData,
      numClasses,
      categoricaFeaturesInfo,
      numTrees,
      featureSubsetStrategy,
      impurity,
      maxDepth,
      maxBins
    )

    testData
      .map((point: LabeledPoint) => model.predict(point.features))
      .take(3)
      .foreach(println)

    spark.stop()
  }
}

堆栈跟踪的顶部如下：

java.io.IOException: unexpected exception type
    at java.io.ObjectStreamClass.throwMiscException(ObjectStreamClass.java:1736)
    at java.io.ObjectStreamClass.invokeReadResolve(ObjectStreamClass.java:1266)
    at java.io.ObjectInputStream.readOrdinaryObject(ObjectInputStream.java:2078)
    at java.io.ObjectInputStream.readObject0(ObjectInputStream.java:1573)
    at java.io.ObjectInputStream.defaultReadFields(ObjectInputStream.java:2287)
    at java.io.ObjectInputStream.readSerialData(ObjectInputStream.java:2211)
    at java.io.ObjectInputStream.readOrdinaryObject(ObjectInputStream.java:2069)
    at java.io.ObjectInputStream.readObject0(ObjectInputStream.java:1573)
    at java.io.ObjectInputStream.defaultReadFields(ObjectInputStream.java:2287)
    at java.io.ObjectInputStream.readSerialData(ObjectInputStream.java:2211)
    at java.io.ObjectInputStream.readOrdinaryObject(ObjectInputStream.java:2069)
    at java.io.ObjectInputStream.readObject0(ObjectInputStream.java:1573)
    at java.io.ObjectInputStream.defaultReadFields(ObjectInputStream.java:2287)
    at java.io.ObjectInputStream.readSerialData(ObjectInputStream.java:2211)
    at java.io.ObjectInputStream.readOrdinaryObject(ObjectInputStream.java:2069)
    at java.io.ObjectInputStream.readObject0(ObjectInputStream.java:1573)
    at java.io.ObjectInputStream.readObject(ObjectInputStream.java:431)
    at org.apache.spark.serializer.JavaDeserializationStream.readObject(JavaSerializer.scala:75)
    at org.apache.spark.serializer.JavaSerializerInstance.deserialize(JavaSerializer.scala:114)
    at org.apache.spark.scheduler.ResultTask.runTask(ResultTask.scala:83)
    at org.apache.spark.scheduler.Task.run(Task.scala:121)
    at org.apache.spark.executor.Executor$TaskRunner$$anonfun$10.apply(Executor.scala:402)
    at org.apache.spark.util.Utils$.tryWithSafeFinally(Utils.scala:1360)
    at org.apache.spark.executor.Executor$TaskRunner.run(Executor.scala:408)
    at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
    at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
    at java.lang.Thread.run(Thread.java:748)
Caused by: java.lang.reflect.InvocationTargetException
    at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
    at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62)
    at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
    at java.lang.reflect.Method.invoke(Method.java:498)
    at java.lang.invoke.SerializedLambda.readResolve(SerializedLambda.java:230)
    at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
    at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62)
    at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
    at java.lang.reflect.Method.invoke(Method.java:498)
    at java.io.ObjectStreamClass.invokeReadResolve(ObjectStreamClass.java:1260)
    ... 25 more
Caused by: java.lang.BootstrapMethodError: java.lang.NoClassDefFoundError: scala/runtime/LambdaDeserialize
    at ItsABreeze$.$deserializeLambda$(ItsABreeze.scala)
    ... 35 more
Caused by: java.lang.NoClassDefFoundError: scala/runtime/LambdaDeserialize
    ... 36 more
Caused by: java.lang.ClassNotFoundException: scala.runtime.LambdaDeserialize
    at java.net.URLClassLoader.findClass(URLClassLoader.java:382)
    at java.lang.ClassLoader.loadClass(ClassLoader.java:424)
    at java.lang.ClassLoader.loadClass(ClassLoader.java:357)

Answer 1

我尝试运行的代码是此页面上 classification example 的略微修改版本（来自 Spark 机器学习库文档）。

我最初的问题的两位评论者都是正确的：我将我使用的 Scala 版本从 2.12.8 更改为 2.11.12，并将 Spark 恢复为 2.2.1，并且代码运行正常。

对于任何有资格回答这个问题的人，这里有一个后续问题：Spark 2.4.0 声称对 Scala 2.12.x 有新的实验性支持。 2.12.x 支持是否存在很多已知问题？