DanielLeens commented on issue #11499: URL: https://github.com/apache/seatunnel/issues/11499#issuecomment-5040696773
Thanks for opening #11516 and for carrying the analysis through the extra Parquet split audit. The root cause still looks exactly as described earlier: the Parquet path drops `tableId` when it creates `FileSourceSplit`, and Spark later blows up when the internal table-id field is serialized as null. The added note that the row-group and split-size strategies already preserve `tableId` is useful because it keeps the fix narrowly targeted instead of over-editing unrelated split paths. This issue should stay open until #11516 lands. Once it is merged, it would be good to validate one real Hive parquet table again on Spark, just to confirm the regression test matches the runtime path the reporter hit. -- 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]
