[jira] [Created] (FLINK-11276) Slicing Window Optimization

[Shang Yuanchun (Jira)]

Rong Rong created FLINK-11276:
---------------------------------

             Summary: Slicing Window Optimization
                 Key: FLINK-11276
                 URL: https://issues.apache.org/jira/browse/FLINK-11276
             Project: Flink
          Issue Type: New Feature
            Reporter: Rong Rong


This is the umbrella ticket for the discussion in:
https://docs.google.com/document/d/1ziVsuW_HQnvJr_4a9yKwx_LEnhVkdlde2Z5l6sx5HlY/edit?usp=sharing

We would like to support syntax for more flexible window aggregations:
1. slicing and merging:

{code:java}
val resultStream: DataStream = inputStream
  .keyBy("key")
  .sliceWindow(Time.seconds(5L))       // new “slice window” concept: to combine
                                       // tumble results based on discrete
                                       //  non-overlapping windows.
  .sum("value")
  .slideOver(Count.of(5))              // new “slide over” concept to merge
                                       // tumble results.
{code}

and
2. sliding aggregation with multi intervals, such as:

{code:java}
val slicedStream: SlicedStream = inputStream
  .keyBy("key")
  .sliceWindow(Time.seconds(5L))
  .sum("value")
val resultStream1: DataStream = slicedStream
  .slideOver(Time.seconds(10L))        // slide over 5-sec slices
                                       // for the last 10 seconds
val resultStream2: DataStream = slicedStream
  .slideOver(Count.of(3))              // slide over 5-sec slices
                                       // for the last 3 slices (15 seconds)
val resultStream3: DataStream = slicedStream
  .slideOver(Count.of(180))            // slide over 5-sec slices
                                       // for the last 3 slices (15 minutes)
{code}

and
3. replace current sliding window logic with more efficient slice and merge approach.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)