Generally speaking, a window defines a finite set of elements on an unbounded stream. This set can be based on time, element counts, a combination of counts and time, or some custom logic to assign elements to windows.
- Flink documentation (flink.apache.org)
Dealing with infinite data stream demands the need for such window functions. Flink’s DataStream API (discussed in following sections) does have some built-in windowing functions that takes care of most use cases. It also allows us to define custom window behavior as required by your use case by letting developers implement its interfaces and implementing appropriate methods.
The following are Flink’s built-in windowing options:
- Time window
- Count window