Hi Jia! This sounds a bit fishy. The docs mention that there is no need for a timestamp / watermark extractor because with 0.10, the timestamps that come with Kafka records can be used directly to produce watermarks for event time.
One quick clarification: did you also check whether the timestamps that come with the Kafka 0.10 records are sound and reasonable? Cheers, Gordon On 17 May 2017 at 4:45:19 AM, Jia Teoh (jiat...@gmail.com) wrote: Hi, I'm trying to use KafkaConsumer010 as a source for a windowing job on event time, as provided by Kafka. According to the kafka connector doc (link), I've set the time characteristic to event time (streamExecutionEnvironment.setStreamTimeCharacteristic(TimeCharacteristic.EventTime)) and am using KafkaConsumer010 along with Kafka 0.10.2. I've also set up windowing: "stream.windowAll(TumblingEventTimeWindows.of(Time.milliseconds(windowMillis)))" (using timeWindowAll appears to be equivalent as well). Using these configurations I can verify that data is read from Kafka. However, the event time windows never trigger even when data is loaded for much longer than the window size. Is there an additional configuration I am missing? I have verified that the Kafka messages have timestamps. The docs mention that there is no need for a timestamp extractor, but using one to explicitly assign the current time does result in windows being triggered. Thanks, Jia Teoh
