Flink does not support per key watermarks or type sensitive watermarks. The underlying assumption is that you have a global watermark which defines the progress wrt to event time in your topology.
The easiest way would be to have an input which has a monotonically increasing timestamp. Alternatively you can define the maximum lag between the watermark and the timestamp and then generate watermarks with w = timestamp - maxLag. That way you allow elements to be out of order for a certain amount of event time. Cheers, Till On Tue, Nov 8, 2016 at 5:02 PM, Sendoh <unicorn.bana...@gmail.com> wrote: > Thank you for confirming. > > What would you think an efficient way not having global watermark? The > following logic fails to build Watermark per KeyStream: > jsonStreams.keyBy(new JsonKeySelector()).assignTimestampsAndWatermarks(new > JsonWatermark()).keyBy(JsonKeySelector()).window(.... > > So, using split(), or implementing an event type recognized > AssignerWithPeriodicWatermarks along with custom EventTimeTrigger would be > the solution? > > Best, > > Sendoh > > > > -- > View this message in context: http://apache-flink-user- > mailing-list-archive.2336050.n4.nabble.com/Cannot-see-all- > events-in-window-apply-for-big-input-tp9945p9988.html > Sent from the Apache Flink User Mailing List archive. mailing list archive > at Nabble.com. >
