andygrove commented on PR #1962: URL: https://github.com/apache/datafusion-ballista/pull/1962#issuecomment-4905067953
Verified this fix on a 2-executor cluster (2×8) reading TPC-H SF100 Parquet from MinIO over `s3://`, with the adaptive planner enabled (`-c ballista.planner.adaptive.enabled=true`). Before this change, AQE jobs against a remote object store failed at planning with `No suitable object store found`. With this branch (built from `bug_1950`), the same queries plan and execute successfully: | Query (AQE + S3) | Result | |---|---| | Q1 | 6.35 s, 4 rows | | Q5 | 41.8 s, 5 rows | | Q9 | 46.8 s, 175 rows | Confirmed AQE was actually engaged (not a static-planner fallback): the scheduler logged the `Adaptive Query Planning is EXPERIMENTAL` warning per job and the plans contain `AdaptiveDatafusionExec` nodes. **Zero** `No suitable object store found` errors across the run, and row counts are correct. LGTM 👍 -- 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]
