spark serialization issue with hashmap

Question

I am trying to serialize a Java class whose objects will be collected in Java Spark RDD. The POJO contains some data structures like HashMap and ArrayList.

When I am trying to iterate over the RDD objects in my driver and try to get the hashmap elements , its throwing a Serialization exception.

java.io.IOException: com.esotericsoftware.kryo.KryoException: java.lang.NullPointerException

Sample Program :

Driver Class:

Main Implementaion :

public class MyMainTest {
    public static ArrayList<MyInterface> call() throws Exception{
        ArrayList<MyInterface> li = new ArrayList<MyInterface>();
        for(int i=0;i<3;i++){
            MyRecord  myrec = new MyRecord();
            li.add(myrec.addToTagMap());
        }
        return li;
    }
}

Main Implementation:
public class MyMainTest {
    public static ArrayList<MyInterface> call() throws Exception{
        ArrayList<MyInterface> li = new ArrayList<MyInterface>();
        for(int i=0;i<3;i++){
            MyRecord  myrec = new MyRecord();
            li.add(myrec.addToTagMap());
        }
        return li;
    }
}

My Pojo Class:
public class MyRecord implements MyInterface,Serializable{


    protected HashMap<String, Object> allTagNameValueMap_
    =  new HashMap<String, Object>(){
        {
            put("key1", "T1");
            put("key2", "Val2");
            put("key3", 3);
        }};
        protected MyRecord addToTagMap() throws Exception  {
            MyRecord myRec = new MyRecord();
            myRec.allTagNameValueMap_.put("key1", "New Value");
            myRec.allTagNameValueMap_.put("key2", "New Value2");
            myRec.allTagNameValueMap_.put("key4", 22);

            return  myRec;
        }
        @Override
        public Object getKey1() {
            // TODO Auto-generated method stub
            return allTagNameValueMap_.get("key1");
        }
        @Override
        public Object getKey2() {
            // TODO Auto-generated method stub
            return allTagNameValueMap_.get("key2");
        }
        @Override
        public Object getKey3() {
            // TODO Auto-generated method stub
            return allTagNameValueMap_.get("key3");
        }
}

My Interface:
public interface MyInterface extends Serializable{

    public Object getKey1();
    public Object getKey2();
    public Object getKey3();
}

Serialization trace:

allTagNameValueMap_ (com.org.util.ex.ExRecord)

at org.apache.spark.util.Utils$.tryOrIOException(Utils.scala:1140)
at org.apache.spark.rdd.ParallelCollectionPartition.readObject(ParallelCollectionRDD.scala:70)
at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:57)
at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
at java.lang.reflect.Method.invoke(Method.java:606)
at java.io.ObjectStreamClass.invokeReadObject(ObjectStreamClass.java:1017)
at java.io.ObjectInputStream.readSerialData(ObjectInputStream.java:1893)
at java.io.ObjectInputStream.readOrdinaryObject(ObjectInputStream.java:1798)
at java.io.ObjectInputStream.readObject0(ObjectInputStream.java:1350)
at java.io.ObjectInputStream.defaultReadFields(ObjectInputStream.java:1990)
at java.io.ObjectInputStream.readSerialData(ObjectInputStream.java:1915)
at java.io.ObjectInputStream.readOrdinaryObject(ObjectInputStream.java:1798)
at java.io.ObjectInputStream.readObject0(ObjectInputStream.java:1350)
at java.io.ObjectInputStream.readObject(ObjectInputStream.java:370)
at org.apache.spark.serializer.JavaDeserializationStream.readObject(JavaSerializer.scala:68)
at org.apache.spark.serializer.JavaSerializerInstance.deserialize(JavaSerializer.scala:94)
at org.apache.spark.executor.Executor$TaskRunner.run(Executor.scala:185)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1145)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:615)
at java.lang.Thread.run(Thread.java:745)

Caused by: com.esotericsoftware.kryo.KryoException: java.lang.NullPointerException Serialization trace:

allTagNameValueMap_ (com.org.util.ex.ExRecord)

at com.esotericsoftware.kryo.serializers.FieldSerializer$ObjectField.read(FieldSerializer.java:626)
at com.esotericsoftware.kryo.serializers.FieldSerializer.read(FieldSerializer.java:221)
at com.esotericsoftware.kryo.Kryo.readClassAndObject(Kryo.java:729)
at com.esotericsoftware.kryo.serializers.DefaultArraySerializers$ObjectArraySerializer.read(DefaultArraySerializers.java:338)
at com.esotericsoftware.kryo.serializers.DefaultArraySerializers$ObjectArraySerializer.read(DefaultArraySerializers.java:293)
at com.esotericsoftware.kryo.Kryo.readClassAndObject(Kryo.java:729)
at com.twitter.chill.WrappedArraySerializer.read(WrappedArraySerializer.scala:36)
at com.twitter.chill.WrappedArraySerializer.read(WrappedArraySerializer.scala:23)
at com.esotericsoftware.kryo.Kryo.readClassAndObject(Kryo.java:729)
at org.apache.spark.serializer.KryoDeserializationStream.readObject(KryoSerializer.scala:138)
at org.apache.spark.rdd.ParallelCollectionPartition$$anonfun$readObject$1$$anonfun$apply$mcV$sp$2.apply(ParallelCollectionRDD.scala:80)
at org.apache.spark.rdd.ParallelCollectionPartition$$anonfun$readObject$1$$anonfun$apply$mcV$sp$2.apply(ParallelCollectionRDD.scala:80)
at org.apache.spark.util.Utils$.deserializeViaNestedStream(Utils.scala:130)
at org.apache.spark.rdd.ParallelCollectionPartition$$anonfun$readObject$1.apply$mcV$sp(ParallelCollectionRDD.scala:80)
at org.apache.spark.util.Utils$.tryOrIOException(Utils.scala:1137)
... 20 more

Caused by: java.lang.NullPointerException at java.util.HashMap.put(HashMap.java:493) at com.esotericsoftware.kryo.serializers.MapSerializer.read(MapSerializer.java:135) at com.esotericsoftware.kryo.serializers.MapSerializer.read(MapSerializer.java:17) at com.esotericsoftware.kryo.Kryo.readObject(Kryo.java:648) at com.esotericsoftware.kryo.serializers.FieldSerializer$ObjectField.read(FieldSerializer.java:605)

Is there any particular things to be noted while serializing HashMap in Spark?

Welcome to SO Hari. Could your provide a [MCVE](http://stackoverflow.com/help/mcve)? It would be easier to help you and more useful for further readers. In a meantime check [this](http://stackoverflow.com/a/33042316/1560062). — zero323, Oct 15 '15 at 19:27
Thanks . zero323 Should I make a call to MyKryoRegistrator from my Driver Program and then set the sparkConf serializable as KryoSerializable — Harisyam, Oct 15 '15 at 23:37
I don't want to sound rude, but I really suggest you to try Scala, you will see how your code will look shorter and will be easier to understand! I am not saying your code is bad or anything! just in case — Alberto Bonsanto, Oct 15 '15 at 23:52
Yes , Scala is really good . But I have some limitations and right now has to stick on with Java . — Harisyam, Oct 16 '15 at 01:39

score 0 · Accepted Answer · answered Oct 22 '15 at 00:04

I fixed the issue .

The steps I followed .

Create a Kryo registrator class and register the Pojo as Kryo Serializable .

public class MyKryoRegistrator implements KryoRegistrator, Serializable {
  @Override
  public void registerClasses(Kryo kryo) {
      // Product POJO associated to a product Row from the DataFrame      
      kryo.register(MyRecord.class);  
  }
}

2.Set the serializable type and the Kryo registrator class in the spark conf inside my main.

sConf.set("spark.serializer","org.apache.spark.serializer.KryoSerializer");
sConf.set("spark.kryo.registrator","com.spark.util.umf.MyKryoRegistrator");

3.Initialize the HashMap using a constructor in the POJO class .

protected Map allTagNameValueMap_ =new HashMap(); public MyRecord(){ allTagNameValueMap_.put("key1", "V1"); allTagNameValueMap_.put("key2", "V2"); allTagNameValueMap_.put("key3", "V3"); };

spark serialization issue with hashmap

1 Answers1