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

Posted to dev@flink.apache.org by Yufei Zhang <af...@apache.org> on 2022/08/24 05:18:16 UTC

[upsert-kafka][SQL] Why Upsert-Kafka SQL connector is a different connector from the Kafka SQL Connector

Hi all,

I noticed there is a Upsert-Kafka SQL Connector[1] and the Kafka SQL
Connector[2], my question is why do we implement them as two different
connectors? For example, why not add a `upsert` config option in the Kafka
SQL connector to specify whether the table enables upsert mode or not. Is
there a special consideration that we have to use a different connector
name like `upsert-kafka` and implement a different
DynamicSource/SinkFactory  ?

Thank you in advance~

Cheers,
Yufei







- [1] :
https://nightlies.apache.org/flink/flink-docs-release-1.15/docs/connectors/table/upsert-kafka/

- [2]:
https://nightlies.apache.org/flink/flink-docs-release-1.15/docs/connectors/table/kafka

Re: [upsert-kafka][SQL] Why Upsert-Kafka SQL connector is a different connector from the Kafka SQL Connector

Posted by Yufei Zhang <af...@gmail.com>.

Thanks~ Will take a look.

Cheers,
Yufei

On Wed, Aug 24, 2022 at 3:45 PM Jark Wu <im...@gmail.com> wrote:

> There was a long discussion about this. You can read the thread[1].
>
> Best,
> Jark
>
> [1]: https://lists.apache.org/thread/sr6on0cq52ns4y3zh6j0x899by80n5zw
>
> On Wed, 24 Aug 2022 at 13:20, Yufei Zhang <af...@apache.org> wrote:
>
> > Hi all,
> >
> > I noticed there is a Upsert-Kafka SQL Connector[1] and the Kafka SQL
> > Connector[2], my question is why do we implement them as two different
> > connectors? For example, why not add a `upsert` config option in the
> Kafka
> > SQL connector to specify whether the table enables upsert mode or not. Is
> > there a special consideration that we have to use a different connector
> > name like `upsert-kafka` and implement a different
> > DynamicSource/SinkFactory  ?
> >
> > Thank you in advance~
> >
> > Cheers,
> > Yufei
> >
> >
> >
> >
> >
> >
> >
> > - [1] :
> >
> >
> https://nightlies.apache.org/flink/flink-docs-release-1.15/docs/connectors/table/upsert-kafka/
> >
> > - [2]:
> >
> >
> https://nightlies.apache.org/flink/flink-docs-release-1.15/docs/connectors/table/kafka
> >
>

Re: [upsert-kafka][SQL] Why Upsert-Kafka SQL connector is a different connector from the Kafka SQL Connector

Posted by Jark Wu <im...@gmail.com>.

There was a long discussion about this. You can read the thread[1].

Best,
Jark

[1]: https://lists.apache.org/thread/sr6on0cq52ns4y3zh6j0x899by80n5zw

On Wed, 24 Aug 2022 at 13:20, Yufei Zhang <af...@apache.org> wrote:

> Hi all,
>
> I noticed there is a Upsert-Kafka SQL Connector[1] and the Kafka SQL
> Connector[2], my question is why do we implement them as two different
> connectors? For example, why not add a `upsert` config option in the Kafka
> SQL connector to specify whether the table enables upsert mode or not. Is
> there a special consideration that we have to use a different connector
> name like `upsert-kafka` and implement a different
> DynamicSource/SinkFactory  ?
>
> Thank you in advance~
>
> Cheers,
> Yufei
>
>
>
>
>
>
>
> - [1] :
>
> https://nightlies.apache.org/flink/flink-docs-release-1.15/docs/connectors/table/upsert-kafka/
>
> - [2]:
>
> https://nightlies.apache.org/flink/flink-docs-release-1.15/docs/connectors/table/kafka
>