Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
RELOAD, a novel learned query optimizer, tackles the instability and inefficiency of existing RL-based optimizers by focusing on minimizing query-level performance regressions and accelerating convergence to expert-level plan quality. It achieves this through a combination of techniques including a novel reward shaping mechanism and a robust exploration strategy. Empirical results on standard benchmarks show RELOAD achieves up to 2.4x higher robustness and 3.1x greater efficiency compared to state-of-the-art RL-based query optimization techniques.
RL-based query optimizers can finally be robust and efficient enough for production, outperforming state-of-the-art methods by up to 2.4x in robustness and 3.1x in efficiency.
Recent advances in query optimization have shifted from traditional rule-based and cost-based techniques towards machine learning-driven approaches. Among these, reinforcement learning (RL) has attracted significant attention due to its ability to optimize long-term performance by learning policies over query planning. However, existing RL-based query optimizers often exhibit unstable performance at the level of individual queries, including severe performance regressions, and require prolonged training to reach the plan quality of expert, cost-based optimizers. These shortcomings make learned query optimizers difficult to deploy in practice and remain a major barrier to their adoption in production database systems. To address these challenges, we present RELOAD, a robust and efficient learned query optimizer for database systems. RELOAD focuses on (i) robustness, by minimizing query-level performance regressions and ensuring consistent optimization behavior across executions, and (ii) efficiency, by accelerating convergence to expert-level plan quality. Through extensive experiments on standard benchmarks, including Join Order Benchmark, TPC-DS, and Star Schema Benchmark, RELOAD demonstrates up to 2.4x higher robustness and 3.1x greater efficiency compared to state-of-the-art RL-based query optimization techniques.