terrymanu commented on issue #37276:
URL:
https://github.com/apache/shardingsphere/issues/37276#issuecomment-3651510684
Sorry, after thinking it through, we shouldn’t silently truncate. Updated
change request:
- Production logic: Don’t align columns in
UnicastResourceShowExecutor#createQueryResult by truncating/padding. Find the
root cause of the header/data mismatch and fix QueryResponseHeader at the
source (build from the real ResultSetMetaData, and branch by database type/
version if needed) so data.size() and columnCount already match, rather
than trimming afterward.
- Transparency: If you still need protection against out-of-bounds, throw
a clear exception or emit an observable warning; avoid silently dropping data.
- Likely mismatch sources (for investigation): backend version/dialect
differences changing SHOW/system view column sets (e.g., MySQL 8.0 vs MariaDB);
the proxy rewrites/filters headers but doesn’t filter row data; metadata
templates/caches lag actual view changes;
driver/protocol returns hidden/extended columns ignored by header
construction; heterogeneous shards (different versions/dialects) returning
different column counts during merge.
--
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]
