I3eka commented on PR #45561:
URL: https://github.com/apache/superset/pull/45561#issuecomment-6415688025

   @rebenitez1802 Addressed all seven points from [your 
review](https://github.com/apache/superset/pull/45561#pullrequestreview-5605869932)
 in 8698a1afe6:
   
   - Cache failures now log a warning and preserve the successful result and 
stored Query state.
   - Both error and success paths preserve an accepted Stop and close the 
transaction when the guard returns False.
   - MCP worker deadlines finalize an existing Query as timed out before 
re-raising the original BaseException signal. The shared exception has moved to 
the SQL cancellation module, keeping optional MCP imports out of the executor.
   - A failed Query INSERT now recovers the inactive session even when the 
caller has no Query object yet.
   - Failure finalization preserves an existing end time.
   - Tests share the metadata-session and error fixtures, no longer use 
`create=True`, and check persisted state and transaction health instead of 
exact commit/rollback counts.
   
   The six behavior findings reproduced in eight targeted scenarios before the 
fixes. Final validation: **868 passed, 3 skipped** across SQL execution, MCP 
worker, and execute_sql tests; another **8 passed** with `--log-level=ERROR`. 
Changed-file pre-commit checks pass. SQLite verifies transaction boundaries and 
persisted outcomes; live Postgres/MySQL row-lock behavior was not tested.
   
   I also replied to every inline review thread, separating confirmed defects 
from the annotation/refresh reports that the current code already handles. The 
PR description and MCP deadline documentation are updated. Ready for another 
look.
   


-- 
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]

Reply via email to