how aggregate total sum in a period of time(window) with kafka?

The code below is only process the first message arrived and publishes it properly. But no more messages are processed afterwards(I'm using kafka-console-consumer.bat in the terminal to monitor the messages published to total-amount-by-id)

Kafka Streams:

KStream<String, String> totalAmount = builder.stream("data-consumed", Consumed.with(Serdes.String(), Serdes.String()));


totalAmount
  .mapValues(v -> Integer.valueOf(v))
  .groupByKey()
  .windowedBy(TimeWindows.of(Duration.ofMillis(100)))
  .aggregate(
        () -> new Integer(0),
        (key, value, aggregate) -> {
                        System.out.println("value: "+value);
                        System.out.println("aggregate: "+aggregate);
                        return value+aggregate;
                    },,
        Materialized.with(Serdes.String(), Serdes.Integer())
  )
  .toStream()
  .map(((key, aggregate) -> new KeyValue<>(key.key(), aggregate)))
  .to("total-amount-by-id", Produced.with(Serdes.String(), Serdes.Integer()));

Tests:

• I'm publishing 1 message per 100ms, always with the same key, to the topic "data-consumed"
• The first four (k,v) published to the topic "data-consumed" were: (1,1),(1,2),(1,4),(1,1)
• Kafka streams published (1,1) to "total-amount-by-id", but nothing else came after that
• The System.out.println() in the code above printed only: value: 1 aggregate: 0 value: 2 aggregate: 0

Any guess of the reason behind this problem?

*I was expecting to have the second aggregate equals to 1(aggregate: 1)