[
https://issues.apache.org/jira/browse/FLINK-37156?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17916558#comment-17916558
]
cheng qian commented on FLINK-37156:
------------------------------------
Hi yux,
Thank you for your reply. I have reviewed FLINK-36794 and the related
implementation, but it's still a bit different from my scenario. The objective
of FLINK-36794 is to have multiple source instances within a single Pipeline.
My idea, on the other hand, is to have one source corresponding to multiple
sinks in a Pipeline. The aim is to reduce repeated data collection and send a
single set of data to multiple different downstream systems.
> A Flink pipeline supports collecting data once and writing it to multiple
> sinks.
> --------------------------------------------------------------------------------
>
> Key: FLINK-37156
> URL: https://issues.apache.org/jira/browse/FLINK-37156
> Project: Flink
> Issue Type: New Feature
> Affects Versions: cdc-3.3.0
> Reporter: cheng qian
> Priority: Major
>
> Is there any plan in the community to support the scenario of multiple sinks?
> We want to synchronize data from a MySQL database to StarRocks, Paimon, and
> Kafka. Currently, we need to create three Pipelines, which triples the I/O
> load on the MySQL database. We envision defining three sinks within a single
> Pipeline, enabling data to be collected once and written to three different
> downstream systems.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
