[
https://issues.apache.org/jira/browse/FLINK-24289?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17415364#comment-17415364
]
Jark Wu commented on FLINK-24289:
---------------------------------
Please use `StatementSet` to combine two DMLs into one job, then Flink can
helps to only just load source table once.
see
https://nightlies.apache.org/flink/flink-docs-master/docs/dev/table/common/#translate-and-execute-a-query
> Load data repeatedly between different jobs
> -------------------------------------------
>
> Key: FLINK-24289
> URL: https://issues.apache.org/jira/browse/FLINK-24289
> Project: Flink
> Issue Type: Improvement
> Components: Table SQL / API
> Affects Versions: 1.12.4
> Reporter: Spongebob
> Priority: Major
>
> I had created an jdbc source table named tbl_a, and then I executed two DMLs
> that both visited tbl_a in await mode, it means flink executed DML1 and then
> DML2. Out of my expetation that flink had load tbl_a from underlying database
> both in DML1 and DML2.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
