Open
Description
I'm integrating Cube.js with Trino and encountering performance issues due to the way Cube.js handles queries. Specifically, Cube.js appends an "ORDER BY" clause to all queries directed to Trino. This becomes a bottleneck because, even with a LIMIT clause (e.g., LIMIT 10), Trino performs a full dataset scan when an "ORDER BY" is present, leading to significant slowdowns when dealing with large tables. I connect to Cube.js using psql. Is there an approach to either disable the automatic "ORDER BY" addition or to optimize query performance under these conditions?"