具有自定义对象数据类型的 Kafka 流聚合

Question

我有一个处理器，它从主题中获取 json 字符串，类型为 GenericRecord。 现在我将流分成 2 个分支。 我采用第一个分支并将 (key, value) 映射到 2 个字符串中，其中包含 json 的特定字段和该字段的值，并且我按键分组。 到目前为止一切都很好。 现在我必须使用用户定义的新类型聚合流，并且收到异常。

代码如下：

新类型：

private class Tuple {

    public int occ;
    public int sum;


    public Tuple (int occ, int sum) {
        this.occ = occ;
        this.sum = sum;
    }

    public void sum (int toAdd) {
        this.sum += toAdd;
        this.occ ++;
    }

    public int getAverage () {
        return this.sum / this.occ;
    }

    public String toString() {
        return occ + "-> " + sum + ": " + getAverage();
    }

好流：

  StreamsBuilder builder = new StreamsBuilder();
    KStream<GenericRecord, GenericRecord> source =
          builder.stream(topic);

    KStream<GenericRecord, GenericRecord>[] branches = source.branch(
            (key,value) -> partition(value.toString()),
            (key, value) -> true
    );

    KGroupedStream <String, String> groupedStream = branches[0]
            .mapValues(value -> createJson(value.toString()))
            .map((key, value) -> KeyValue.pair(new String("T_DUR_CICLO"), value.getNumberValue("payload", "T_DUR_CICLO")))
            .peek((key, value) -> System.out.println("key=" + key + ", value=" + value))
            .groupByKey();

问题：

   KTable<String, Tuple> aggregatedStream = groupedStream.aggregate(
            () -> new Tuple (0,0), // initializer 
            (aggKey, newValue, aggValue) -> new Tuple (aggValue.occ + 1, aggValue.sum + Integer.parseInt(newValue)));



    KafkaStreams streams = new KafkaStreams(builder.build(), props);
    streams.start();

这是个例外：

   Exception in thread "streamtest-6173d6a2-4a3a-4d76-b793-774719f8b1f5-StreamThread-1" org.apache.kafka.streams.errors.StreamsException: Exception caught in process. taskId=1_0, processor=KSTREAM-SOURCE-0000000011, topic=streamtest-KSTREAM-AGGREGATE-STATE-STORE-0000000007-repartition, partition=0, offset=0
    at org.apache.kafka.streams.processor.internals.StreamTask.process(StreamTask.java:318)
    at org.apache.kafka.streams.processor.internals.AssignedStreamsTasks.process(AssignedStreamsTasks.java:94)
    at org.apache.kafka.streams.processor.internals.TaskManager.process(TaskManager.java:409)
    at org.apache.kafka.streams.processor.internals.StreamThread.processAndMaybeCommit(StreamThread.java:964)
    at org.apache.kafka.streams.processor.internals.StreamThread.runOnce(StreamThread.java:832)
    at org.apache.kafka.streams.processor.internals.StreamThread.runLoop(StreamThread.java:767)
    at org.apache.kafka.streams.processor.internals.StreamThread.run(StreamThread.java:736)
Caused by: org.apache.kafka.streams.errors.StreamsException: A serializer (value: io.confluent.kafka.streams.serdes.avro.GenericAvroSerializer) is not compatible to the actual value type (value type: com.mycompany.maventest.Streamer$Tuple). Change the default Serdes in StreamConfig or provide correct Serdes via method parameters.
    at org.apache.kafka.streams.state.StateSerdes.rawValue(StateSerdes.java:195)
    at org.apache.kafka.streams.state.internals.MeteredKeyValueBytesStore$1.innerValue(MeteredKeyValueBytesStore.java:66)
    at org.apache.kafka.streams.state.internals.MeteredKeyValueBytesStore$1.innerValue(MeteredKeyValueBytesStore.java:57)
    at org.apache.kafka.streams.state.internals.InnerMeteredKeyValueStore.put(InnerMeteredKeyValueStore.java:206)
    at org.apache.kafka.streams.state.internals.MeteredKeyValueBytesStore.put(MeteredKeyValueBytesStore.java:117)
    at org.apache.kafka.streams.kstream.internals.KStreamAggregate$KStreamAggregateProcessor.process(KStreamAggregate.java:94)
    at org.apache.kafka.streams.processor.internals.ProcessorNode$1.run(ProcessorNode.java:50)
    at org.apache.kafka.streams.processor.internals.ProcessorNode.runAndMeasureLatency(ProcessorNode.java:244)
    at org.apache.kafka.streams.processor.internals.ProcessorNode.process(ProcessorNode.java:133)
    at org.apache.kafka.streams.processor.internals.ProcessorContextImpl.forward(ProcessorContextImpl.java:143)
    at org.apache.kafka.streams.processor.internals.ProcessorContextImpl.forward(ProcessorContextImpl.java:126)
    at org.apache.kafka.streams.processor.internals.ProcessorContextImpl.forward(ProcessorContextImpl.java:90)
    at org.apache.kafka.streams.processor.internals.SourceNode.process(SourceNode.java:87)
    at org.apache.kafka.streams.processor.internals.StreamTask.process(StreamTask.java:302)
    ... 6 more
Caused by: java.lang.ClassCastException: com.mycompany.maventest.Streamer$Tuple cannot be cast to org.apache.avro.generic.GenericRecord
    at io.confluent.kafka.streams.serdes.avro.GenericAvroSerializer.serialize(GenericAvroSerializer.java:39)
    at org.apache.kafka.streams.state.StateSerdes.rawValue(StateSerdes.java:191)
    ... 19 more

我该如何解决这个问题？

----- 更新 ------

生产者使用 Avro 生产，所以我有这个配置属性：

 props.put(StreamsConfig.DEFAULT_KEY_SERDE_CLASS_CONFIG, GenericAvroSerde.class);
 props.put(StreamsConfig.DEFAULT_VALUE_SERDE_CLASS_CONFIG, GenericAvroSerde.class);

如果我指定了自定义 serde，结果如下：

 KTable<String, Tuple> aggregatedStream = groupedStream.aggregate(
            () -> new Tuple(0, 0), // initializer 
            (aggKey, newValue, aggValue) ->  new Tuple (aggValue.occ + 1, aggValue.sum + Integer.parseInt(newValue)),
            Materialized.with(Serdes.String(), new MySerde()));

例外：

   Exception in thread "streamtest-17deb5c8-ed07-4fcf-bd59-37b75e44b83f-StreamThread-1" org.apache.kafka.streams.errors.StreamsException: Deserialization exception handler is set to fail upon a deserialization error. If you would rather have the streaming pipeline continue after a deserialization error, please set the default.deserialization.exception.handler appropriately.
    at org.apache.kafka.streams.processor.internals.RecordDeserializer.deserialize(RecordDeserializer.java:80)
    at org.apache.kafka.streams.processor.internals.RecordQueue.addRawRecords(RecordQueue.java:97)
    at org.apache.kafka.streams.processor.internals.PartitionGroup.addRawRecords(PartitionGroup.java:117)
    at org.apache.kafka.streams.processor.internals.StreamTask.addRecords(StreamTask.java:677)
    at org.apache.kafka.streams.processor.internals.StreamThread.addRecordsToTasks(StreamThread.java:943)
    at org.apache.kafka.streams.processor.internals.StreamThread.runOnce(StreamThread.java:831)
    at org.apache.kafka.streams.processor.internals.StreamThread.runLoop(StreamThread.java:767)
    at org.apache.kafka.streams.processor.internals.StreamThread.run(StreamThread.java:736)
Caused by: org.apache.kafka.common.errors.SerializationException: Error deserializing Avro message for id -1
Caused by: org.apache.kafka.common.errors.SerializationException: Unknown magic byte!

--- 已解决 ---- 我还为 groupBy 中的类型更改添加了新的 serde

 KGroupedStream <String, String> groupedStream = branches[0]
            .mapValues(value -> createJson(value.toString()))
            .map((key, value) -> KeyValue.pair(new String("T_DUR_CICLO"), value.getNumberValue("payload", "T_DUR_CICLO")))
            .peek((key, value) -> System.out.println("key=" + key + ", value=" + value))
            .groupByKey( Serialized.with(
                    Serdes.String(), /* key (note: type was modified) */
                    Serdes.String()));  /* value */

Answer 1

Kafka 流将使用默认 Serde，除非在操作中明确指定。

在 aggregate() 方法中，您将 valueType 定义为 Tuple 而默认 serde 是 GenericRecord 因此它会引发异常。您需要指定 serde，如下所示：

 KTable<String, Tuple> aggregatedStream = groupedStream.aggregate(
            () -> new Tuple (0,0), // initializer 
            (aggKey, newValue, aggValue) -> 
                 new Tuple (aggValue.occ + 1, aggValue.sum + Integer.parseInt(newValue))
                ,Materialized.with(keySerde, tupleSerde));

它将使用 tupleSerde 进行此操作。您可以在此处找到示例： https://docs.confluent.io/current/streams/developer-guide/dsl-api.html#aggregating