[
https://issues.apache.org/jira/browse/FLINK-32613?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Timo Walther closed FLINK-32613.
--------------------------------
Fix Version/s: 1.18.0
Resolution: Fixed
Fixed in master: 663a68f2c56fe6bcdb50fb2c07229cffc3e948e6
> Disable check for single rowtime attribute for sinks
> ----------------------------------------------------
>
> Key: FLINK-32613
> URL: https://issues.apache.org/jira/browse/FLINK-32613
> Project: Flink
> Issue Type: Improvement
> Components: Table SQL / Planner
> Reporter: Timo Walther
> Assignee: Timo Walther
> Priority: Major
> Labels: pull-request-available
> Fix For: 1.18.0
>
>
> A common error that users face is the following:
> {code}
> The query contains more than one rowtime attribute column [%s] for writing
> into table '%s'.
> Please select the column that should be used as the event-time timestamp for
> the
> table sink by casting all other columns to regular TIMESTAMP or TIMESTAMP_LTZ.
> {code}
> However, not every sink requires the rowtime set on the stream record. For
> example, the Kafka sink does not use it because it exposes metadata columns.
> The user can define which column is the rowtime column by projection.
> Either we introduce a config option for disabling this check. Or we come up
> with an interface that a connector can implement. I would vote for the config
> option as an easy solution to get rid of this error message.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
