zhangyue19921010 opened a new issue, #11257: URL: https://github.com/apache/paimon/issues/11257
### Search before asking - [x] I searched in the [issues](https://github.com/apache/paimon/issues) and found nothing similar. ### Paimon version Master `79b8ec4788e2cd8ed87b1745d5ca54aa8db0e293`. ### Compute Engine Java Table API, OpenJDK 11.0.18. ### Minimal reproduce step 1. Create a primary-key table with `k INT`, `v INT`, primary key `k`, `bucket=1`, and `deletion-vectors.enabled=true`. 2. Write `(1, 10)` and `(2, 20)`, commit with compaction awaited, and save the current snapshot ID. 3. Delete `(1, 10)` and commit with compaction awaited. A full read now correctly returns only `(2, 20)`. 4. Read the audit-log table with `incremental-between=<saved snapshot>,<latest snapshot>` and `incremental-between-scan-mode=diff`. The diff is empty, although it should contain `-D(1, 10)`. The same workflow without deletion vectors returns the deletion. Updating a key to the same value also incorrectly emits a change in DV mode. ### What doesn't meet your expectations? Snapshot diff should compare visible rows at the two endpoints. The planner removes identical data-file manifest entries from both sides even when their deletion vectors differ, so the reader never sees the changed visibility of those rows. ### Anything else? For primary-key diffs, retain a shared file when its before/after deletion-file descriptors differ. Continue pruning files with unchanged descriptors. Append-table diff planning retains its existing semantics. ### Are you willing to submit a PR? - [x] I'm willing to submit a PR! -- 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]
