Hi Jing,
    Thanks for your response. As stated in the FLIP, the purpose of this
FLIP is meant to support
user-defined async table function. As described in flink document [1]

Async table functions are special functions for table sources that perform
> a lookup.
>

So end user can not directly define and use async table function now. An
user case is reported in [2]

So, in conclusion, no new interface is introduced, but we extend the
ability to support user-defined async table function.

[1]:
https://nightlies.apache.org/flink/flink-docs-release-1.17/docs/dev/table/functions/udfs/
[2]: https://lists.apache.org/thread/qljwd40v5ntz6733cwcdr8s4z97b343b

Thanks.
Aitozi.


Jing Ge <j...@ververica.com.invalid> 于2023年5月27日周六 06:40写道:

> Hi Aitozi,
>
> Thanks for your proposal. I am not quite sure if I understood your thoughts
> correctly. You described a special case implementation of the
> AsyncTableFunction with on public API changes. Would you please elaborate
> your purpose of writing a FLIP according to the FLIP documentation[1]?
> Thanks!
>
> [1]
>
> https://cwiki.apache.org/confluence/display/FLINK/Flink+Improvement+Proposals
>
> Best regards,
> Jing
>
> On Wed, May 24, 2023 at 1:07 PM Aitozi <gjying1...@gmail.com> wrote:
>
> > May I ask for some feedback  :D
> >
> > Thanks,
> > Aitozi
> >
> > Aitozi <gjying1...@gmail.com> 于2023年5月23日周二 19:14写道:
> > >
> > > Just catch an user case report from Giannis Polyzos for this usage:
> > >
> > > https://lists.apache.org/thread/qljwd40v5ntz6733cwcdr8s4z97b343b
> > >
> > > Aitozi <gjying1...@gmail.com> 于2023年5月23日周二 17:45写道:
> > > >
> > > > Hi guys,
> > > >     I want to bring up a discussion about adding support of User
> > > > Defined AsyncTableFunction in Flink.
> > > > Currently, async table function are special functions for table
> source
> > > > to perform
> > > > async lookup. However, it's worth to support the user defined async
> > > > table function.
> > > > Because, in this way, the end SQL user can leverage it to perform the
> > > > async operation
> > > > which is useful to maximum the system throughput especially for IO
> > > > bottleneck case.
> > > >
> > > > You can find some more detail in [1].
> > > >
> > > > Looking forward to feedback
> > > >
> > > >
> > > > [1]:
> >
> https://cwiki.apache.org/confluence/display/FLINK/%5BFLIP-313%5D+Add+support+of+User+Defined+AsyncTableFunction
> > > >
> > > > Thanks,
> > > > Aitozi.
> >
>

Reply via email to