[ https://issues.apache.org/jira/browse/FLINK-21007?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17267056#comment-17267056 ]
Dawid Wysakowicz edited comment on FLINK-21007 at 1/18/21, 7:53 AM: -------------------------------------------------------------------- Please use English in JIRA. If you intend to ask a question in Chinese you can use the user...@flink.apache.org mailing [list|https://flink.apache.org/community.html]. was (Author: dawidwys): Please use English in JIRA. If you intend to ask a question in Chinese you can use the user...@flink.apache.org mailing [list|https://flink.apache.org/community.html. > flink sql sink 支持更新目标表的部分字段以及提供null值处理选项 > ---------------------------------------- > > Key: FLINK-21007 > URL: https://issues.apache.org/jira/browse/FLINK-21007 > Project: Flink > Issue Type: Improvement > Components: Connectors / JDBC > Affects Versions: 1.12.2 > Reporter: sky > Priority: Blocker > Fix For: 1.13.0 > > > 在通过flink SQL作为etl使用的时候,需要flink sql sink > 支持更新目标表的部分字段以及提供null值处理选项(1、当前record剔除null值的字段,只更新没有null的字段到目标表,2、即使结果中有null的字段依旧更新到目标表)。 > eg: > 目标表 > -------------------------- > id | name | age | level| > -------------------------- > 1 | sky1 | 30 | 1 | > -------------------------- > 3 | sky3 | 30 | 2 | > -------------------------- > 计算完写入存储的records(比目标表少一列): > {"id":1,level:2} > {"id":2,level:1} > 计算结果只想根据id更新level字段,不能用null覆盖name和age字段,并且需要考虑id=2的数据行不存在的可能。 -- This message was sent by Atlassian Jira (v8.3.4#803005)