繁体   English   中英

火花提交中的 java.lang.NoSuchMethodError

[英]java.lang.NoSuchMethodError in spark-submit

使用sbt package编译我的代码并在 spark 中提交它们后:

sudo -u spark spark-submit  --master yarn --deploy-mode client --executor-memory 2G --num-executors 6 --class viterbiAlgorithm.viterbiAlgo ./target/scala-2.11/vibertialgo_2.11-1.3.4.jar

我收到了这个错误:

Exception in thread "main" java.lang.NoSuchMethodError: breeze.linalg.DenseVector$.tabulate$mDc$sp(ILscala/Function1;Lscala/reflect/ClassTag;)Lbreeze/linalg/DenseVector;
    at viterbiAlgorithm.User$$anonfun$eval$2.apply(viterbiAlgo.scala:84)
    at viterbiAlgorithm.User$$anonfun$eval$2.apply(viterbiAlgo.scala:80)
    at scala.collection.immutable.Range.foreach(Range.scala:160)
    at viterbiAlgorithm.User.eval(viterbiAlgo.scala:80)
    at viterbiAlgorithm.viterbiAlgo$.main(viterbiAlgo.scala:28)
    at viterbiAlgorithm.viterbiAlgo.main(viterbiAlgo.scala)
    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 org.apache.spark.deploy.JavaMainApplication.start(SparkApplication.scala:52)
    at org.apache.spark.deploy.SparkSubmit.org$apache$spark$deploy$SparkSubmit$$runMain(SparkSubmit.scala:851)
    at org.apache.spark.deploy.SparkSubmit.doRunMain$1(SparkSubmit.scala:167)
    at org.apache.spark.deploy.SparkSubmit.submit(SparkSubmit.scala:195)
    at org.apache.spark.deploy.SparkSubmit.doSubmit(SparkSubmit.scala:86)
    at org.apache.spark.deploy.SparkSubmit$$anon$2.doSubmit(SparkSubmit.scala:926)
    at org.apache.spark.deploy.SparkSubmit$.main(SparkSubmit.scala:935)
    at org.apache.spark.deploy.SparkSubmit.main(SparkSubmit.scala)

sbt 构建文件如下:

name := "vibertiAlgo"  
version := "1.3.4"  
scalaVersion := "2.11.2"  

libraryDependencies  ++= Seq(  
    "org.scalanlp" %% "breeze" % "1.0",  
    "org.apache.spark" %% "spark-core" % "2.4.0",  
    "org.apache.spark" %% "spark-sql" % "2.4.0")

尽管使用sbt run我可以在本地成功运行代码,所以我不认为我的代码有任何问题。 另外,scala和spark的编译和运行版本是一样的。

viterbiAlgo.scala的代码是:

package viterbiAlgorithm

import breeze.linalg._
// import org.apache.spark.sql.SparkSession


object viterbiAlgo {
  def main(arg: Array[String]) {

    val A = DenseMatrix((0.5,0.2,0.3), 
      (0.3,0.5,0.2),
      (0.2,0.3,0.5))
    val B = DenseMatrix((0.5,0.5), 
      (0.4,0.6),
      (0.7,0.3))
    val pi = DenseVector(0.2,0.4,0.4) 

    val o = DenseVector[Int](0,1,0) //Hive time + cell_id
    val model = new Model(A,B,pi) 
    val user = new User("Jack", model, o) //Hive 
    user.eval() // run algorithm
    user.printResult()

    //spark sql
    // val warehouseLocation = "spark-warehouse"
    // val spark = SparkSession.builder().appName("Spark.sql.warehouse.dir").config("spark.sql.warehouse.dir", warehouseLocation).enableHiveSupport().getOrCreate()
    // import spark.implicits._
    // import spark.sql

    // val usr = "1"
    // val model = new Model(A,B,pi) 
    // val get_statement = "SELECT * FROM viterbi.observation"
    // val df = sql(get_statement)
    // val o = DenseVector(df.filter(df("usr")===usr).select(df("obs")).collect().map(_.getInt(0)))
    // val user = new User(usr, model, o)
    // user.eval()
    // user.printResult()
  }
}


class Model (val A: DenseMatrix[Double], val B:DenseMatrix[Double], val pi: DenseVector[Double]) {
  def info():Unit = {
    println("The model is:")
    println("A:")
    println(A)
    println("B:")
    println(B)
    println("Pi:")
    println(pi)
  }
}

class User (val usr_name: String, val model: Model, val o:DenseVector[Int]) {
  val N = model.A.rows // state number
  val M = model.B.cols // observation state
  val T = o.length // time 
  val delta = DenseMatrix.zeros[Double](N,T)
  val psi = DenseMatrix.zeros[Int](N,T)
  val best_route = DenseVector.zeros[Int](T)

  def eval():Unit = {
    //1. Initialization
    delta(::,0) := model.pi * model.B(::, o(0))
    psi(::,0) := DenseVector.zeros[Int](N)

    /*2. Induction
    */
    val tempDelta = DenseMatrix.zeros[Double](N,N)// Initialization
    val tempB = DenseMatrix.zeros[Double](N,N)// Initialization
    for (t <- 1 to T-1) { 
      // Delta
      tempDelta := DenseMatrix.tabulate(N, N){case (i, j) => delta(i,t-1)}
      tempB := DenseMatrix.tabulate(N, N){case (i, j) => model.B(j, o(t))} 
      delta(::, t) := DenseVector.tabulate(N){i => max((tempDelta *:* model.A *:* tempB).t.t(::,i))} 
    }

    //3. Maximum
    val P_star = max(delta(::, T-1))
    val i_star_T = argmax(delta(::, T-1))
    best_route(T-1) = i_star_T

    //4. Backward

    for (t <- T-2 to 0 by -1) {
      best_route(t) = psi(best_route(t+1),t+1)
    }
  }

  def printResult():Unit = {
    println("User: " + usr_name)
    model.info()
    println
    println("Observed: ")
    printRoute(o)
    println("Best_route is: ")
    printRoute(best_route)
    println("delta is")
    println(delta)
    println("psi is: ")
    println(psi)
  }

  def printRoute(v: DenseVector[Int]):Unit = {
    for (i <- v(0 to -2)){
      print(i + "->")
    }
    println(v(-1))
  }

}


我也尝试--jars参数并传递了微风库的位置,但得到了同样的错误。

我需要提到的是,我在服务器上“本地”测试了代码,并在 spark-shell 上测试了所有方法(我可以在服务器上的 spark-shell 上导入微风库)。

服务器 scala 版本与 sbt 构建文件中的版本匹配。 虽然火花版本是 2.4.0-cdh6.2.1,但如果我在“2.4.0”之后添加“cdh6.2.1”,sbt 将无法编译。

我尝试了 Victor 提供的两种可能的解决方案,但没有成功。 但是,我将 sbt 构建文件中的微风版本从1.0更改为0.13.2 ,一切正常。 但我不知道出了什么问题。

如果您在本地而不是在服务器中运行代码,这意味着您没有在您提交的作业的类路径中提供库。

你有两个选择:

  • 使用--jars参数并传递所有库的位置(在您的情况下,它似乎是breeze库)。
  • 使用sbt assembly插件,该插件将生成一个带有所有所需依赖项的胖 JAR,然后将该 JAR 提交给作业。

暂无
暂无

声明:本站的技术帖子网页,遵循CC BY-SA 4.0协议,如果您需要转载,请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.

 
粤ICP备18138465号  © 2020-2024 STACKOOM.COM