Hi Navneeth,
Your understanding is correct.
In the image, all windows across the keys for the same timespan are grouped
together, which make sense from a logical perspective as you would talk
about the first, second, ... window.
But technically, there are 15 small windows involved instead of the
depicted 5 windows. I guess, for you, the technical representation would be
easier to understand.
For session windows, for example, only the technical view makes sense. So
you could also look at examples of these windows if it's easier for you.
On Tue, May 12, 2020 at 10:05 AM Navneeth Krishnan
wrote:
> Hi All,
>
> I was looking at the documentation for windows and got a little confused.
> As per my understanding tumbling window per key will create a non
> overlapping window based on when the data for that key arrived. For example
> consider a tumbling window of 30 seconds
> user1 - 10:01:01
> user2 - 10:01:02
> user1 - 10:01:05
> user2 - 10:01:06
> user2 - 10:01:08
>
> Result:
> user1 (10:01:01 & 10:01:05)
> user2 (10:01:02 & 10:01:06)
> user2 (10:01:08)
>
> Is this the right understanding?
>
> But as per the below image in docs, it looks like the window is not per
> key, please correct me if i'm wrong.
> [image: image.png]
>
> Thanks
>
--
Arvid Heise | Senior Java Developer
<https://www.ververica.com/>
Follow us @VervericaData
--
Join Flink Forward <https://flink-forward.org/> - The Apache Flink
Conference
Stream Processing | Event Driven | Real Time
--
Ververica GmbH | Invalidenstrasse 115, 10115 Berlin, Germany
--
Ververica GmbH
Registered at Amtsgericht Charlottenburg: HRB 158244 B
Managing Directors: Timothy Alexander Steinert, Yip Park Tung Jason, Ji
(Toni) Cheng
