cshuo opened a new pull request, #20481: URL: https://github.com/apache/hudi/pull/20481
### Describe the issue this Pull Request addresses Closes #20480 Flink incremental batch and streaming reads can prune partitions containing historical records using current partition statistics. After an update narrows a partition's range, a historical query may miss matching records. ### Summary and Changelog - Avoid constructing the partition stats pruner for incremental batch reads and streaming reads, while retaining partition-field pruning. - Cover historical incremental reads for COW and MOR tables, batch and streaming modes, and the legacy source and Source V2. - Verify snapshot and read-optimized reads retain partition stats pruning. ### Impact Incremental reads no longer use partition stats pruning, which may increase the number of partitions scanned. Snapshot and read-optimized batch reads retain existing behavior. ### Risk Level Low. This disables an optimization for incremental reads while preserving partition-field pruning. Targeted Flink 2.2 tests cover historical reads and pruning behavior. ### Documentation Update None. ### Contributor's checklist - [ ] Read through [contributor's guide](https://hudi.apache.org/contribute/how-to-contribute) - [x] Enough context is provided in the sections above - [x] Adequate tests were added if applicable -- 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]
