trino-query-optimization

What problem does it solve?

It helps you diagnose and speed up slow Trino distributed SQL queries by improving filtering, projection, join strategy, dynamic filtering, and execution efficiency so the engine does less work and moves less data.

Core Features & Use Cases

• Filter and projection pushdown: ensures predicates and selected columns are applied as early as possible (e.g., avoiding full scans and SELECT *).
• Join optimization and dynamic filtering: chooses broadcast vs partitioned joins, improves join ordering via CBO/ANALYZE, and leverages runtime dynamic filters to reduce scanned data.
• Shuffle and session tuning: minimizes expensive repartitioning and adjusts session properties (memory, exchange compression, task parallelism) to reduce bottlenecks.

Quick Start

Ask the agent to optimize your slow Trino SQL by checking the execution plan for pushdown failures, inefficient projections, poor join strategy, missing partition pruning, excessive shuffles, and stale statistics, then propose revised SQL and relevant EXPLAIN/SKIIP checks.