You are viewing a plain text version of this content. The canonical link for it is here.

Posted to issues@flink.apache.org by "Fabian Hueske (JIRA)" <ji...@apache.org> on 2018/04/04 11:46:00 UTC

[jira] [Commented] (FLINK-9066) Add 'HOTSPOTS' function to Flink SQL

    [ https://issues.apache.org/jira/browse/FLINK-9066?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16425369#comment-16425369 ] 

Fabian Hueske commented on FLINK-9066:
--------------------------------------

This definitely looks interesting!

We'd have to check what kind of function that would be and whether this type of functions and the syntax are already supported by Calcite.
The internals of the function are not explained in detail by the post, but it might be some kind of syntactic sugar around an OVER window aggregation (which is already supported). 

We would need to come up with a generic approach to detect these hotspots.

> Add 'HOTSPOTS' function to Flink SQL
> ------------------------------------
>
>                 Key: FLINK-9066
>                 URL: https://issues.apache.org/jira/browse/FLINK-9066
>             Project: Flink
>          Issue Type: Improvement
>          Components: Table API &amp; SQL
>            Reporter: Bowen Li
>            Priority: Major
>
> Inspired by ['HOTSPOTS' function in AWS Kinesis Analytics SQL|https://aws.amazon.com/blogs/aws/real-time-hotspot-detection-in-amazon-kinesis-analytics/].
> cc [~fhueske] [~fhueske@gmail.com]



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