ggjh-159 commented on issue #12459: URL: https://github.com/apache/gluten/issues/12459#issuecomment-4953905452
> [@ggjh-159](https://github.com/ggjh-159) Thanks for the detailed . I have a question about the RowKind propagation issue. > > The description says that RowKind is dropped during the `RowData`→ `RowVector` conversion, so Velox no longer receives any RowKind information. > > However, it also says: > > `As a result, the velox side sees every row as +I regardless of the original changelog marker.` > > Could you clarify why Velox sees every row as +I specifically? > > If the RowKind is completely lost, where does the +I come from? Is INSERT (+I) the default `RowKind` when reconstructing `RowData` in `Flink` ( GenericRowData), or is it being assumed somewhere else in the Gluten-Flink pipeline? @Moemenmohamed24 Since there is no RowKind representation in Velox, all data is treated as raw data, which is implicitly treated as Flink's INSERT state data. -- 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] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
