wwj6591812 commented on PR #8117: URL: https://github.com/apache/paimon/pull/8117#issuecomment-4644984637
> The validation here still allows `scan.bucket` on non-fixed-bucket tables. > > `validateSpecifiedBucket` rejects postpone bucket tables, but it does not require a fixed bucket mode or `bucket > 0`. For bucket-unaware/dynamic bucket tables, `CoreOptions.bucket()` can be `<= 0`, so the upper-bound check is skipped and the scan proceeds with physical bucket pruning. That can turn an invalid configuration into an empty/incorrect result instead of a clear error. The generated doc says this option is only supported for fixed-bucket primary-key tables (`bucket > 0`). > > Could we enforce that here, e.g. require the fixed bucket mode and configured bucket count > 0, and also check primary-key-ness if the option is intended only for primary-key tables? Hi, thanks for your review. I have updated code, now alidateSpecifiedBucket requires HASH_FIXED bucket mode, primary keys, and bucket > 0. Added core UT (ReadBuilderImplTest) and extended Flink UT/IT for invalid table types. Please CC, Thx. @JingsongLi -- 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]
