CalvinKirs commented on PR #67960: URL: https://github.com/apache/doris/pull/67960#issuecomment-5677270738
This reintroduces the exact bug #63292 fixed. BE still sends TIMESTAMPTZ as a string with the `+08:00` suffix in both protocols (`MysqlRowBuffer::push_timestamptz` -> `push_string`); `MysqlResultSender.sendBinaryResultRow` only covers FE-local result sets. So with the column advertised as TIMESTAMP, Connector/J fails with the same "Invalid length (32) for type TIMESTAMP" on server-prepared statements, `getTimestamp` fails on the text protocol too, and go-sql-driver with `parseTime=true` rejects the value with "invalid time bytes". #66953 is not fixed, and `test_timestamptz_jdbc_binary_protocol.groovy` would have shown you this. On the bigger issue: 14 auto-generated PRs in two days across FE, BE, CDC, Paimon and CI, none of them verified against the existing tests or a real client. That is not contributing, it is pushing your verification onto maintainers. I'm not reviewing the rest of the batch. If you want one looked at, run the existing tests, reproduce the bug, show it fixed with a real driver, and write the description yourself after reading your own diff. -- 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]
