Standalone spark cluster. Can't submit job programmatically -> java.io.InvalidClassException

Question

Spark fellows, I’m quite new to Spark, that’s why I hope for your help indeed.

I’m trying to schedule the quite simple job on spark cluster from my laptop. Despite it works, when I submit it with ./spark-submit, it throws an exception, when I try to do it programmatically.

Environment: - Spark - 1 master node and 2 worker nodes (standalone mode). Spark was not compiled, but the binaries were downloaded. Spark version - 1.0.2 - java version "1.7.0_45" - Application jar is located everywhere (on client and on worker nodes in the same place); - README.md file is copied to every node as well;

The application I'm trying to run:

val logFile = "/user/vagrant/README.md"

val conf = new SparkConf()
conf.setMaster("spark://192.168.33.50:7077")
conf.setAppName("Simple App")
conf.setJars(List("file:///user/vagrant/spark-1.0.2-bin-hadoop1/bin/hello-apache-spark_2.10-1.0.0-SNAPSHOT.jar"))
conf.setSparkHome("/user/vagrant/spark-1.0.2-bin-hadoop1")

val sc = new SparkContext(conf)

val logData = sc.textFile(logFile, 2).cache()

...

So the problem is, that this application runs on cluster successfully, when I do:

./spark-submit --class com.paycasso.SimpleApp --master spark://192.168.33.50:7077 --deploy-mode client file:///home/vagrant/spark-1.0.2-bin-hadoop1/bin/hello-apache-spark_2.10-1.0.0-SNAPSHOT.jar

But it doesn't work, when I try to do the same programmatically by calling sbt run

Here is the stacktrace, that I get on master node:

14/09/04 15:09:44 ERROR Remoting: org.apache.spark.deploy.ApplicationDescription; local class incompatible: stream classdesc serialVersionUID = -6451051318873184044, local class serialVersionUID = 583745679236071411
java.io.InvalidClassException: org.apache.spark.deploy.ApplicationDescription; local class incompatible: stream classdesc serialVersionUID = -6451051318873184044, local class serialVersionUID = 583745679236071411
    at java.io.ObjectStreamClass.initNonProxy(ObjectStreamClass.java:617)
    at java.io.ObjectInputStream.readNonProxyDesc(ObjectInputStream.java:1622)
    at java.io.ObjectInputStream.readClassDesc(ObjectInputStream.java:1517)
    at java.io.ObjectInputStream.readOrdinaryObject(ObjectInputStream.java:1771)
    at java.io.ObjectInputStream.readObject0(ObjectInputStream.java:1350)
    at java.io.ObjectInputStream.defaultReadFields(ObjectInputStream.java:1990)
    at java.io.ObjectInputStream.readSerialData(ObjectInputStream.java:1915)
    at java.io.ObjectInputStream.readOrdinaryObject(ObjectInputStream.java:1798)
    at java.io.ObjectInputStream.readObject0(ObjectInputStream.java:1350)
    at java.io.ObjectInputStream.readObject(ObjectInputStream.java:370)
    at akka.serialization.JavaSerializer$$anonfun$1.apply(Serializer.scala:136)
    at scala.util.DynamicVariable.withValue(DynamicVariable.scala:57)
    at akka.serialization.JavaSerializer.fromBinary(Serializer.scala:136)
    at akka.serialization.Serialization$$anonfun$deserialize$1.apply(Serialization.scala:104)
    at scala.util.Try$.apply(Try.scala:161)
    at akka.serialization.Serialization.deserialize(Serialization.scala:98)
    at akka.remote.serialization.MessageContainerSerializer.fromBinary(MessageContainerSerializer.scala:58)
    at akka.serialization.Serialization$$anonfun$deserialize$1.apply(Serialization.scala:104)
    at scala.util.Try$.apply(Try.scala:161)
    at akka.serialization.Serialization.deserialize(Serialization.scala:98)
    at akka.remote.MessageSerializer$.deserialize(MessageSerializer.scala:23)
    at akka.remote.DefaultMessageDispatcher.payload$lzycompute$1(Endpoint.scala:55)
    at akka.remote.DefaultMessageDispatcher.payload$1(Endpoint.scala:55)
    at akka.remote.DefaultMessageDispatcher.dispatch(Endpoint.scala:73)
    at akka.remote.EndpointReader$$anonfun$receive$2.applyOrElse(Endpoint.scala:764)
    at akka.actor.ActorCell.receiveMessage(ActorCell.scala:498)
    at akka.actor.ActorCell.invoke(ActorCell.scala:456)
    at akka.dispatch.Mailbox.processMailbox(Mailbox.scala:237)
    at akka.dispatch.Mailbox.run(Mailbox.scala:219)
    at akka.dispatch.ForkJoinExecutorConfigurator$AkkaForkJoinTask.exec(AbstractDispatcher.scala:386)
    at scala.concurrent.forkjoin.ForkJoinTask.doExec(ForkJoinTask.java:260)
    at scala.concurrent.forkjoin.ForkJoinPool$WorkQueue.runTask(ForkJoinPool.java:1339)
    at scala.concurrent.forkjoin.ForkJoinPool.runWorker(ForkJoinPool.java:1979)
    at scala.concurrent.forkjoin.ForkJoinWorkerThread.run(ForkJoinWorkerThread.java:107)

What could be the solution to this? Thank you in advance.

Answer 1

If you are using a prebuilt spark. (No sbt or maven install), make sure that all the worker nodes use the same version of spark. I faced the same issue as one of worker nodes was using a different version of spark. keep the same versions across all nodes solved the issue for me. One machine was using spark-2.0.0--bin-hadoop2.7 instead of spark-2.0.0-preview-bin-hadoop2.7

Answer 2

Instead of using the regular Spark Core library, you might use such dependency (defined on the build.sbt file):

resolvers += "Talend" at "https://talend-update.talend.com/nexus/content/repositories/libraries/"
libraryDependencies += "org.talend.libraries" % "spark-assembly-1.6.0-hadoop2.6.0" % "6.0.0"

/!\ Those spark-assembly-...hadoop... librairies can be quite large (and therefore not compatible with a Git push).

List of Spark/Hadoop Libraries Hosted by Talend

Answer 3

After wasting a lot of time, I've found the problem. Despite I haven't used hadoop/hdfs in my application, hadoop client matters. The problem was in hadoop-client version, it was different than the version of hadoop, spark was built for. Spark's hadoop version 1.2.1, but in my application that was 2.4.

When I changed the version of hadoop client to 1.2.1 in my app, I'm able to execute spark code on cluster.