Spark: object not serializable - scala

Question

I have a batch job which I am try to convert to structured streaming. I am getting the following error:
20/03/31 15:09:23 ERROR Executor: Exception in task 0.0 in stage 0.0 (TID 0)
java.io.NotSerializableException: com.apple.ireporter.analytics.compute.AggregateKey
Serialization stack:
- object not serializable (class: com.apple.ireporter.analytics.compute.AggregateKey, value: d_)
... where "d_" is the last row in the dataset
This is the relevant code snippet
df.writeStream.foreachBatch { (batchDF: DataFrame, batchId: Long) =>
import spark.implicits._
val javaRdd = batchDF.toJavaRDD
val dataframeToRowColFunction = new RowToColumn(table)
println("Back to Main class")
val combinedRdd =javaRdd.flatMapToPair(dataframeToRowColFunction.FlatMapData2).combineByKey(aggrCreateComb.createCombiner,aggrMerge.aggrMerge,aggrMergeCombiner.aggrMergeCombiner)
// spark.createDataFrame( combinedRdd).show(1); // I commented this
// combinedRdd.collect() // I added this as a test
}
This is the FlatMapData2 class
val FlatMapData2: PairFlatMapFunction[Row, AggregateKey, AggregateValue] = new PairFlatMapFunction[Row, AggregateKey, AggregateValue]() {
//val FlatMapData: PairFlatMapFunction[Row, String, AggregateValue] = new PairFlatMapFunction[Row, String, AggregateValue]() {
override def call(x: Row) = {
val tuples = new util.ArrayList[Tuple2[AggregateKey, AggregateValue]]
val decomposedEvents = decomposer.decomposeDistributed(x)
decomposedEvents.foreach {
y => tuples.add(Tuple2(y._1,y._2))
}
tuples.iterator()
}
}
Here is the aggregate Key class
class AggregateKey(var partitionkeys: Map[Int,Any],var clusteringkeys : Map[Int,Any]) extends Comparable [AggregateKey]{
...
}
I am new to this and any help would be appreciated. Please let me know if anything else needs to be added

Answer 1

I was able to solve this problem by making the AggregateKey extend java.io.Serializable
class AggregateKey(var partitionkeys: Map[Int,Any],var clusteringkeys : Map[Int,Any]) extends java.io.Serializable{