为什么在Flink中使用reduce运算符时得到2项

Question

我尝试通过以下代码进行测试以获得具有最小字段频率的Kafka记录，但最后我得到两条记录而不是一条具有最小字段频率的记录。

public static void main(String[] args) throws Exception {

    StreamExecutionEnvironment env = StreamExecutionEnvironment.createLocalEnvironment().setParallelism(1);
    env.getConfig().disableSysoutLogging();

    env.getConfig().setRestartStrategy(RestartStrategies.fixedDelayRestart(4, 10000));
    env.enableCheckpointing(5000); // create a checkpoint every 5 seconds
    env.getConfig().setGlobalJobParameters(parameterTool); // make parameters available in the web interface
    env.setStreamTimeCharacteristic(TimeCharacteristic.EventTime);

    DataStream<Tuple2<String, Tuple1<Integer>>> input = env
            .addSource(
                    new FlinkKafkaConsumer010<>(
                            parameterTool.getRequired("input-topic"),
                            new KafkaEventSchema(),
                            parameterTool.getProperties())
                            .assignTimestampsAndWatermarks(new CustomWatermarkExtractor()))
            .keyBy("word")
           .window(TumblingProcessingTimeWindows.of(Time.seconds(10)))
            .reduce(new MyReduceFunction(), new MyWindowFunction());
    input.print()
    ;

    env.execute("Kafka 0.10 Example");
}

private static class MyReduceFunction implements ReduceFunction<KafkaEvent> {
    public KafkaEvent reduce(KafkaEvent r1, KafkaEvent r2) {
        return r1.getFrequency() > r2.getFrequency() ?
                r2: r1;
    }
}

private static class MyWindowFunction
        implements WindowFunction<KafkaEvent, Tuple2<String, Tuple1<Integer>>, Tuple, TimeWindow> {
    public void apply(Tuple key,
                      TimeWindow window,
                      Iterable<KafkaEvent> minReadings,
                      Collector<Tuple2<String, Tuple1<Integer>>> out) {
        //   Tuple1<Integer> min = minReadings.iterator().next();
        KafkaEvent max = minReadings.iterator().next();
        //out.collect(new Tuple2<Long, SensorReading>(window.getStart(), max));
        out.collect(new Tuple2<String, Tuple1<Integer>>(max.getWord(), new Tuple1(max.getFrequency())));
    }
}



public class KafkaEvent {

    private String word;
    private int frequency;
    private long timestamp;
}

我得到了什么：

(?? t word1111,(1))
(?? t word1111,(11))

我的期望：

(?? t word1111,(1))

KAFKA的测试数据：

    String word = "word1111";
    result = kafkaProducerClientUsage1.sendStringMessageWithByteArray(word, 9);
    result = kafkaProducerClientUsage1.sendStringMessageWithByteArray(word, 5);
    result = kafkaProducerClientUsage1.sendStringMessageWithByteArray(word, 11);
    result = kafkaProducerClientUsage1.sendStringMessageWithByteArray(word, 8);
    result = kafkaProducerClientUsage1.sendStringMessageWithByteArray(word, 1);