I have read about the Window operator <https://ci.apache.org/projects/flink/flink-docs-release-1.13/docs/dev/table/tableapi/#group-windows> in Flink documentation and know that it groups rows into finite groups based on time or row-count intervals.
I saw an example of a sliding count window right there <https://ci.apache.org/projects/flink/flink-docs-release-1.13/docs/dev/table/tableapi/#slide-sliding-windows> : // Sliding Row-count window (assuming a processing-time attribute "proctime").window(Slide.over(rowInterval(10)).every(rowInterval(5)).on($("proctime")).as("w")); As mentioned in the docs, the on method here is to define: > The time attribute to group (time interval) or sort (row count) on. For > batch queries this might be any Long or Timestamp attribute. For streaming > queries this must be a declared event-time or processing-time time > attribute. On the other hand, I searched found this countWindow <https://ci.apache.org/projects/flink/flink-docs-release-1.13/api/java/org/apache/flink/streaming/api/datastream/KeyedStream.html#countWindow-long-long-> method in Flink's Java docs and saw that it does not specify any time-related parameter. I'm wondering why a sliding count window in Flink Table APIs requires processing time whereas it is unnecessary in the Datastream APIs. I really appreciate it if someone can clarify this for me. -- ------------------------------------------------------------ -------------------------------------------------- Nguyen Dich Long, School of Information and Communication Technology (SoICT - https://www.soict.hust.edu.vn) Hanoi University of Science and Technology (https://www.hust.edu.vn) 601, B1 Building - No 1, Dai Co Viet Street, Hai Ba Trung District, Ha Noi, Vietnam Tel: +84 (0)3.54.41.76.76 Email: long.nd162...@sis.hust.edu.vn; longnguyen25111...@gmail.com
