fresh-borzoni opened a new issue, #3443:
URL: https://github.com/apache/fluss/issues/3443

   ### Search before asking
   
   - [x] I searched in the [issues](https://github.com/apache/fluss/issues) and 
found nothing similar.
   
   
   ### Fluss version
   
   main (development)
   
   ### Please describe the bug 🐞
   
   A full primary-key point query on a lake-enabled partitioned table (FULL) is 
rewritten to a kv single-row lookup. It reads only Fluss kv, so for a partition 
expired from Fluss but still in the lake it returns empty, yet the union 
scan(or a partial-key filter) returns the row. 
   
   The COUNT(*) pushdown already guards this via isDataLakeEnabled and the 
single-row lookup doesn't.
   
   ### Solution
   
   Gate the lookup on partition liveness, else fall back to the union scan.
   
   ### 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]

Reply via email to