You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@spark.apache.org by "Attila Zsolt Piros (Jira)" <ji...@apache.org> on 2023/01/31 21:08:00 UTC
[jira] [Commented] (SPARK-42068) Implicit conversion is not working with parallelization in scala with java 11 and spark3
[ https://issues.apache.org/jira/browse/SPARK-42068?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17682756#comment-17682756 ]
Attila Zsolt Piros commented on SPARK-42068:
--------------------------------------------
This issue is because of JDK bugfix ([https://bugs.openjdk.org/browse/JDK-8172726]) which sets the context class loader to the system class loader for fork join pools, for details see [this code|https://github.com/openjdk/jdk/blame/jdk-11%2B28/src/java.base/share/classes/java/util/concurrent/ForkJoinPool.java#L721].
The JDK bugfix is only available from JDK9 so this is why it is not an issue on Java 8.
In case of Java 8 (on a Scala REPL/Spark shell) the context class loader is the same as callers class loader: {*}scala.tools.nsc.interpreter.IMain$TranslatingClassLoader{*}. This is needed as in case of a REPL for each line new code is generated and compiled. This can be only resolved by this TranslatingClassLoader.
You can check this generated code parts by passing *-Xprint:typer* to Scala REPL/spark-shell. Passing this argument reveal where the *$read* object mentioned in the exception ("scala.ScalaReflectionException: object $read not found") is coming from.
*The workaround*
We just need to use a ForkJoinPool which does not changes the context class loader
{noformat}
import java.util.concurrent._
import java.util.concurrent.ForkJoinPool.ForkJoinWorkerThreadFactory
...
class CustomForkJoinWorkerThread(pool: ForkJoinPool) extends ForkJoinWorkerThread(pool)
object customForkJoinWorkerThreadFactory extends ForkJoinWorkerThreadFactory {
override def newThread(pool: ForkJoinPool): ForkJoinWorkerThread =
new CustomForkJoinWorkerThread(pool)
}
val forkJoinPool = new ForkJoinPool(maxTasks, customForkJoinWorkerThreadFactory, null, false)
{noformat}
In case of common pools *ForkJoinPool.commonPool()* one need to set the "java.util.concurrent.ForkJoinPool.common.threadFactory" system property to the new fully-qualified name of the CustomForkJoinWorkerThread but this only have effect to the common pools not those pools which created directly by calling a constructor of the ForkJoinPool.
> Implicit conversion is not working with parallelization in scala with java 11 and spark3
> ----------------------------------------------------------------------------------------
>
> Key: SPARK-42068
> URL: https://issues.apache.org/jira/browse/SPARK-42068
> Project: Spark
> Issue Type: Bug
> Components: Spark Core
> Affects Versions: 3.3.1, 3.2.3, 3.4.0
> Environment: spark version 3.3.1 Using Scala version 2.12.15 (OpenJDK 64-Bit Server VM, Java 11.0.17)
> Reporter: Srinivas Rishindra Pothireddi
> Priority: Major
>
> The following code snippet fails with java 11 with spark3, but works with java 8. It also works with spark2 and java 11.
> {code:java}
> import scala.collection.mutable
> import scala.collection.parallel.{ExecutionContextTaskSupport, ForkJoinTaskSupport}
> case class Person(name: String, age: Int)
> val pc = List(1, 2, 3).par
> val forkJoinPool = new java.util.concurrent.ForkJoinPool(2)
> pc.tasksupport = new ForkJoinTaskSupport(forkJoinPool)
> pc.map { x =>
> val personList: Array[Person] = (1 to 999).map(value => Person("p" + value, value)).toArray
> //creating RDD of Person
> val rddPerson = spark.sparkContext.parallelize(personList, 5)
> val evenAgePerson = rddPerson.filter(_.age % 2 == 0)
> import spark.implicits._
> val evenAgePersonDF = evenAgePerson.toDF("Name", "Age")
> } {code}
> The error is as follows.
> {code:java}
> scala.ScalaReflectionException: object $read not found.
> at scala.reflect.internal.Mirrors$RootsBase.staticModule(Mirrors.scala:185)
> at scala.reflect.internal.Mirrors$RootsBase.staticModule(Mirrors.scala:29)
> at $typecreator6$1.apply(<console>:37)
> at scala.reflect.api.TypeTags$WeakTypeTagImpl.tpe$lzycompute(TypeTags.scala:237)
> at scala.reflect.api.TypeTags$WeakTypeTagImpl.tpe(TypeTags.scala:237)
> at org.apache.spark.sql.catalyst.encoders.ExpressionEncoder$.apply(ExpressionEncoder.scala:52)
> at org.apache.spark.sql.Encoders$.product(Encoders.scala:300)
> at org.apache.spark.sql.LowPrioritySQLImplicits.newProductEncoder(SQLImplicits.scala:261)
> at org.apache.spark.sql.LowPrioritySQLImplicits.newProductEncoder$(SQLImplicits.scala:261)
> at org.apache.spark.sql.SQLImplicits.newProductEncoder(SQLImplicits.scala:32)
> at $anonfun$res0$1(<console>:37)
> at scala.runtime.java8.JFunction1$mcVI$sp.apply(JFunction1$mcVI$sp.java:23)
> at scala.collection.parallel.AugmentedIterableIterator.map2combiner(RemainsIterator.scala:116)
> at scala.collection.parallel.AugmentedIterableIterator.map2combiner$(RemainsIterator.scala:113)
> at scala.collection.parallel.immutable.ParVector$ParVectorIterator.map2combiner(ParVector.scala:66)
> at scala.collection.parallel.ParIterableLike$Map.leaf(ParIterableLike.scala:1064)
> at scala.collection.parallel.Task.$anonfun$tryLeaf$1(Tasks.scala:53)
> at scala.runtime.java8.JFunction0$mcV$sp.apply(JFunction0$mcV$sp.java:23)
> at scala.util.control.Breaks$$anon$1.catchBreak(Breaks.scala:67)
> at scala.collection.parallel.Task.tryLeaf(Tasks.scala:56)
> at scala.collection.parallel.Task.tryLeaf$(Tasks.scala:50)
> at scala.collection.parallel.ParIterableLike$Map.tryLeaf(ParIterableLike.scala:1061)
> at scala.collection.parallel.AdaptiveWorkStealingTasks$WrappedTask.internal(Tasks.scala:160)
> at scala.collection.parallel.AdaptiveWorkStealingTasks$WrappedTask.internal$(Tasks.scala:157)
> at scala.collection.parallel.AdaptiveWorkStealingForkJoinTasks$WrappedTask.internal(Tasks.scala:440)
> at scala.collection.parallel.AdaptiveWorkStealingTasks$WrappedTask.compute(Tasks.scala:150)
> at scala.collection.parallel.AdaptiveWorkStealingTasks$WrappedTask.compute$(Tasks.scala:149)
> at scala.collection.parallel.AdaptiveWorkStealingForkJoinTasks$WrappedTask.compute(Tasks.scala:440)
> at java.base/java.util.concurrent.RecursiveAction.exec(RecursiveAction.java:189)
> at java.base/java.util.concurrent.ForkJoinTask.doExec(ForkJoinTask.java:290)
> at java.base/java.util.concurrent.ForkJoinTask.doJoin(ForkJoinTask.java:396)
> at java.base/java.util.concurrent.ForkJoinTask.join(ForkJoinTask.java:721)
> at scala.collection.parallel.ForkJoinTasks$WrappedTask.sync(Tasks.scala:379)
> at scala.collection.parallel.ForkJoinTasks$WrappedTask.sync$(Tasks.scala:379)
> at scala.collection.parallel.AdaptiveWorkStealingForkJoinTasks$WrappedTask.sync(Tasks.scala:440)
> at scala.collection.parallel.ForkJoinTasks.executeAndWaitResult(Tasks.scala:423)
> at scala.collection.parallel.ForkJoinTasks.executeAndWaitResult$(Tasks.scala:416)
> at scala.collection.parallel.ForkJoinTaskSupport.executeAndWaitResult(TaskSupport.scala:60)
> at scala.collection.parallel.ParIterableLike$ResultMapping.leaf(ParIterableLike.scala:968)
> at scala.collection.parallel.Task.$anonfun$tryLeaf$1(Tasks.scala:53)
> at scala.runtime.java8.JFunction0$mcV$sp.apply(JFunction0$mcV$sp.java:23)
> at scala.util.control.Breaks$$anon$1.catchBreak(Breaks.scala:67)
> at scala.collection.parallel.Task.tryLeaf(Tasks.scala:56)
> at scala.collection.parallel.Task.tryLeaf$(Tasks.scala:50)
> at scala.collection.parallel.ParIterableLike$ResultMapping.tryLeaf(ParIterableLike.scala:963)
> at scala.collection.parallel.AdaptiveWorkStealingTasks$WrappedTask.compute(Tasks.scala:153)
> at scala.collection.parallel.AdaptiveWorkStealingTasks$WrappedTask.compute$(Tasks.scala:149)
> at scala.collection.parallel.AdaptiveWorkStealingForkJoinTasks$WrappedTask.compute(Tasks.scala:440)
> at java.base/java.util.concurrent.RecursiveAction.exec(RecursiveAction.java:189)
> at java.base/java.util.concurrent.ForkJoinTask.doExec(ForkJoinTask.java:290)
> at java.base/java.util.concurrent.ForkJoinPool$WorkQueue.topLevelExec(ForkJoinPool.java:1020)
> at java.base/java.util.concurrent.ForkJoinPool.scan(ForkJoinPool.java:1656)
> at java.base/java.util.concurrent.ForkJoinPool.runWorker(ForkJoinPool.java:1594)
> at java.base/java.util.concurrent.ForkJoinWorkerThread.run(ForkJoinWorkerThread.java:183)
> Suppressed: scala.ScalaReflectionException: object $read not found.
> at scala.reflect.internal.Mirrors$RootsBase.staticModule(Mirrors.scala:185)
> at scala.reflect.internal.Mirrors$RootsBase.staticModule(Mirrors.scala:29)
> at $typecreator6$1.apply(<console>:37)
> at scala.reflect.api.TypeTags$WeakTypeTagImpl.tpe$lzycompute(TypeTags.scala:237)
> at scala.reflect.api.TypeTags$WeakTypeTagImpl.tpe(TypeTags.scala:237)
> at org.apache.spark.sql.catalyst.encoders.ExpressionEncoder$.apply(ExpressionEncoder.scala:52)
> at org.apache.spark.sql.Encoders$.product(Encoders.scala:300)
> at org.apache.spark.sql.LowPrioritySQLImplicits.newProductEncoder(SQLImplicits.scala:261)
> at org.apache.spark.sql.LowPrioritySQLImplicits.newProductEncoder$(SQLImplicits.scala:261)
> at org.apache.spark.sql.SQLImplicits.newProductEncoder(SQLImplicits.scala:32)
> at $anonfun$res0$1(<console>:37)
> at scala.runtime.java8.JFunction1$mcVI$sp.apply(JFunction1$mcVI$sp.java:23)
> at scala.collection.parallel.AugmentedIterableIterator.map2combiner(RemainsIterator.scala:116)
> at scala.collection.parallel.AugmentedIterableIterator.map2combiner$(RemainsIterator.scala:113)
> at scala.collection.parallel.immutable.ParVector$ParVectorIterator.map2combiner(ParVector.scala:66)
> at scala.collection.parallel.ParIterableLike$Map.leaf(ParIterableLike.scala:1064)
> at scala.collection.parallel.Task.$anonfun$tryLeaf$1(Tasks.scala:53)
> at scala.runtime.java8.JFunction0$mcV$sp.apply(JFunction0$mcV$sp.java:23)
> at scala.util.control.Breaks$$anon$1.catchBreak(Breaks.scala:67)
> at scala.collection.parallel.Task.tryLeaf(Tasks.scala:56)
> at scala.collection.parallel.Task.tryLeaf$(Tasks.scala:50)
> at scala.collection.parallel.ParIterableLike$Map.tryLeaf(ParIterableLike.scala:1061)
> at scala.collection.parallel.AdaptiveWorkStealingTasks$WrappedTask.internal(Tasks.scala:160)
> at scala.collection.parallel.AdaptiveWorkStealingTasks$WrappedTask.internal$(Tasks.scala:157)
> at scala.collection.parallel.AdaptiveWorkStealingForkJoinTasks$WrappedTask.internal(Tasks.scala:440)
> at scala.collection.parallel.AdaptiveWorkStealingTasks$WrappedTask.compute(Tasks.scala:150)
> ... 8 more
> Suppressed: scala.ScalaReflectionException: object $read not found.
> at scala.reflect.internal.Mirrors$RootsBase.staticModule(Mirrors.scala:185)
> at scala.reflect.internal.Mirrors$RootsBase.staticModule(Mirrors.scala:29)
> at $typecreator6$1.apply(<console>:37)
> at scala.reflect.api.TypeTags$WeakTypeTagImpl.tpe$lzycompute(TypeTags.scala:237)
> at scala.reflect.api.TypeTags$WeakTypeTagImpl.tpe(TypeTags.scala:237)
> at org.apache.spark.sql.catalyst.encoders.ExpressionEncoder$.apply(ExpressionEncoder.scala:52)
> at org.apache.spark.sql.Encoders$.product(Encoders.scala:300)
> at org.apache.spark.sql.LowPrioritySQLImplicits.newProductEncoder(SQLImplicits.scala:261)
> at org.apache.spark.sql.LowPrioritySQLImplicits.newProductEncoder$(SQLImplicits.scala:261)
> at org.apache.spark.sql.SQLImplicits.newProductEncoder(SQLImplicits.scala:32)
> at $anonfun$res0$1(<console>:37)
> at scala.runtime.java8.JFunction1$mcVI$sp.apply(JFunction1$mcVI$sp.java:23)
> at scala.collection.parallel.AugmentedIterableIterator.map2combiner(RemainsIterator.scala:116)
> at scala.collection.parallel.AugmentedIterableIterator.map2combiner$(RemainsIterator.scala:113)
> at scala.collection.parallel.immutable.ParVector$ParVectorIterator.map2combiner(ParVector.scala:66)
> at scala.collection.parallel.ParIterableLike$Map.leaf(ParIterableLike.scala:1064)
> at scala.collection.parallel.Task.$anonfun$tryLeaf$1(Tasks.scala:53)
> at scala.runtime.java8.JFunction0$mcV$sp.apply(JFunction0$mcV$sp.java:23)
> at scala.util.control.Breaks$$anon$1.catchBreak(Breaks.scala:67)
> at scala.collection.parallel.Task.tryLeaf(Tasks.scala:56)
> at scala.collection.parallel.Task.tryLeaf$(Tasks.scala:50)
> at scala.collection.parallel.ParIterableLike$Map.tryLeaf(ParIterableLike.scala:1061)
> at scala.collection.parallel.AdaptiveWorkStealingTasks$WrappedTask.internal(Tasks.scala:170)
> at scala.collection.parallel.AdaptiveWorkStealingTasks$WrappedTask.internal$(Tasks.scala:157)
> at scala.collection.parallel.AdaptiveWorkStealingForkJoinTasks$WrappedTask.internal(Tasks.scala:440)
> at scala.collection.parallel.AdaptiveWorkStealingTasks$WrappedTask.compute(Tasks.scala:150)
> at scala.collection.parallel.AdaptiveWorkStealingTasks$WrappedTask.compute$(Tasks.scala:149)
> at scala.collection.parallel.AdaptiveWorkStealingForkJoinTasks$WrappedTask.compute(Tasks.scala:440)
> at java.base/java.util.concurrent.RecursiveAction.exec(RecursiveAction.java:189)
> at java.base/java.util.concurrent.ForkJoinTask.doExec(ForkJoinTask.java:290)
> at java.base/java.util.concurrent.ForkJoinPool$WorkQueue.topLevelExec(ForkJoinPool.java:1020)
> at java.base/java.util.concurrent.ForkJoinPool.scan(ForkJoinPool.java:1656)
> at java.base/java.util.concurrent.ForkJoinPool.runWorker(ForkJoinPool.java:1594)
> at java.base/java.util.concurrent.ForkJoinWorkerThread.run(ForkJoinWorkerThread.java:183) {code}
> I noticed the following snippet in the stacktrace.
> {code:java}
> at scala.collection.parallel.Task.$anonfun$tryLeaf$1(Tasks.scala:53)
> at scala.runtime.java8.JFunction0$mcV$sp.apply(JFunction0$mcV$sp.java:23)
> at scala.util.control.Breaks$$anon$1.catchBreak(Breaks.scala:67) {code}
> It seems to be that a java8 runtime is being called even though the environment is java 11.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org
For additional commands, e-mail: issues-help@spark.apache.org