[jira] [Created] (FLINK-12577) Flink uses Kryo serializer if given a SpecificRecord type

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view
|

[jira] [Created] (FLINK-12577) Flink uses Kryo serializer if given a SpecificRecord type

Shang Yuanchun (Jira)
Nico Kruber created FLINK-12577:
-----------------------------------

             Summary: Flink uses Kryo serializer if given a SpecificRecord type
                 Key: FLINK-12577
                 URL: https://issues.apache.org/jira/browse/FLINK-12577
             Project: Flink
          Issue Type: Bug
          Components: API / Type Serialization System
    Affects Versions: 1.8.0, 1.7.2, 1.9.0
            Reporter: Nico Kruber


If a data stream does not return a *subclass* of {{SpecificRecord}}, Flink will actually infer a {{GenericTypeInfo}} and use kryo instead.

Example:
{code}
StreamExecutionEnvironment env = ...
env.getConfig().disableGenericTypes();

DataStream<Measurement> sourceStream = env.addSource(...);
DataStream<SpecificRecord> test = sourceStream.map(value -> new AggregatedSensorStatistics());

public class AggregatedSensorStatistics extends org.apache.avro.specific.SpecificRecordBase implements org.apache.avro.specific.SpecificRecord {...}
{code}

This will lead to
{code}
Exception in thread "main" java.lang.UnsupportedOperationException: Generic types have been disabled in the ExecutionConfig and type org.apache.avro.specific.SpecificRecord is treated as a generic type.
        at org.apache.flink.api.java.typeutils.GenericTypeInfo.createSerializer(GenericTypeInfo.java:86)
        at org.apache.flink.streaming.api.graph.StreamGraph.addOperator(StreamGraph.java:208)
        at org.apache.flink.streaming.api.graph.StreamGraphGenerator.transformOneInputTransform(StreamGraphGenerator.java:541)
        at org.apache.flink.streaming.api.graph.StreamGraphGenerator.transform(StreamGraphGenerator.java:166)
        at org.apache.flink.streaming.api.graph.StreamGraphGenerator.generateInternal(StreamGraphGenerator.java:132)
        at org.apache.flink.streaming.api.graph.StreamGraphGenerator.generate(StreamGraphGenerator.java:124)
        at org.apache.flink.streaming.api.environment.StreamExecutionEnvironment.getStreamGraph(StreamExecutionEnvironment.java:1537)
        at org.apache.flink.streaming.api.environment.LocalStreamEnvironment.execute(LocalStreamEnvironment.java:89)
        at org.apache.flink.streaming.api.environment.StreamExecutionEnvironment.execute(StreamExecutionEnvironment.java:1510)
        at com.ververica.training.statemigration.StateMigrationJobBase.createAndExecuteJob(StateMigrationJobBase.java:68)
        at com.ververica.training.statemigration.avro.StateMigrationJob.main(StateMigrationJob.java:13)
{code}

You may want some flexibility in your types and thus not provide the exact one like {{AggregatedSensorStatistics}} in this example. I don't see any reason we should disallow that behaviour.

Reason for this is that {{TypeExtractor#privateGetForClass()}} is having this code which only sees classes as Avro if they *extend* from {{SpecificRecord}}:
{code}
if (hasSuperclass(clazz, AVRO_SPECIFIC_RECORD_BASE_CLASS)) {
        return AvroUtils.getAvroUtils().createAvroTypeInfo(clazz);
}
{code}



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)