qingfei1994 commented on issue #5079: URL: https://github.com/apache/paimon/issues/5079#issuecomment-2671421133
> [@qingfei1994](https://github.com/qingfei1994) hi, I think that there is no need to validate the existence of `aggregate-functions` during the DDL stage because paimon already support `user-defined aggregate-functions`. > > In our case, tables are managed centrally on Platform A (e.g., for creation and deletion) and only include the basic paimon dependencies. If we enforce the validation of `user-defined aggregate-functions` at the DDL stage, the platform would need to frequently update its paimon dependencies. > > In fact, user-defined aggregate function dependencies only need to be present at runtime. For example, Flink and Spark both support dynamically adding user dependencies. Our case is that we mistakenly config an aggregation function that doesn't exist, and only get exception when we execute the insert. That's why I raised this. What you pointed out makes sense but I'm wondering if we can precheck more as when we use partially update there're validation like this. `Caused by: java.lang.IllegalArgumentException: Must use sequence group for aggregation functions.` -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
