hutiefang76 commented on PR #12516: URL: https://github.com/apache/seatunnel/pull/12516#issuecomment-5876453897
Closing this for now. I rechecked the complete save-mode flow with this PR on its own and found an attached-catalog routing gap. With `RECREATE_SCHEMA` targeting `lake.main.target`, DROP correctly removes the attached table, but the existing metadata lookup and CREATE TABLE path still ignore the catalog. If there is no local table, the replacement is created in local `main`; if a local table already exists, recreation returns without restoring the attached table. I reproduced both cases through `JdbcSaveModeHandler` with DuckDB JDBC 1.3.1.0. The earlier checks covered the individual hooks and the combination with #12506, which missed this standalone failure. This needs to wait for the catalog routing fix and a regression covering the full recreate flow before it is ready for review again. Thanks for taking a look. -- 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]
