DanielLeens commented on issue #9788: URL: https://github.com/apache/seatunnel/issues/9788#issuecomment-5715963909
Thanks @nankeChen75 for preparing a reproduction. **Classification: A / Connector-V2 Doris temporal-value correctness.** I checked current `dev`: Doris `DATETIME`/`DATETIMEV2` is mapped as a local date-time, while the separate `TIMESTAMP_TZ` sink serialization path has explicit timezone handling. The report does not yet identify which boundary changes the value, and there is no matching open PR for this issue or `DATETIMEV2`. Please do not change a shared type converter yet. Before proposing a fix, please add a reproducible text trace containing: 1. The source and sink `CREATE TABLE` DDL, affected column name/type/scale, one exact input value, and the exact stored output value (not screenshots alone). 2. The complete redacted source/sink configuration, SeaTunnel build or `dev` commit, Doris server version, and both worker-JVM and Doris-session time zones. 3. Two small controls that separate the boundary: source Doris -> Console (or an equivalent row capture), then a fixed known SeaTunnel row -> Doris sink using the same target DDL. 4. The result for at least one normal modern date and one value near the observed date, with the same precision and `DATE` versus `DATETIMEV2` type. That evidence will show whether this is source decoding, row serialization, or a Doris-side interpretation issue. Once it isolates one failing boundary, a focused regression test and minimal fix are welcome; please preserve the existing `DATETIMEV2` local-date-time contract outside the proven failure path. -- 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]
