From Blind Search to Memory-Aware Evolution: Efficient DBMS Tuning via Collaborative Diagnosis and Utility-Aware Retrieval

2026-07-20

AI Summary

This work addresses the challenges of multi-component database tuning, which suffers from an exponentially large combinatorial search space and sparse feedback, leading to inefficient existing methods with limited performance gains. To overcome these limitations, we propose EvoTune, a novel framework that jointly diagnoses query-specific high-impact subspaces through synergy between large language model reasoning and lightweight pattern learning, enabling fine-tuning-free subspace identification. EvoTune further introduces a utility-aware memory retrieval mechanism that retrieves historical tuning experiences from a hierarchical memory based on long-term performance gain rather than similarity, facilitating continuous incremental optimization. Under identical tuning budgets, EvoTune achieves up to a 44.5% performance improvement and reaches the final performance of the best baseline up to 3.9× faster.

Abstract

Modern DBMSs expose multiple configurable components (e.g., knobs, query hints, and indexes) that jointly determine query performance. Multi-component tuning is challenging due to the large combinatorial search space and the difficulty of learning effective tuning policies under limited feedback. Existing approaches still rely on blind search over the configuration space and interaction-heavy policy learning, leading to high tuning overhead and limited performance gains. Recent advances in large language models (LLMs) enable knowledge-driven tuning, but existing LLM-based methods fail to effectively exploit online feedback and historical observations, often converging prematurely to suboptimal configurations. In this paper, we present EvoTune, a memory-aware evolution framework for multi-component DBMS tuning. EvoTune first localizes a query-specific high-impact subspace via collaborative diagnosis, which combines lightweight pattern learning with LLM-based reasoning. It further introduces a utility-aware retrieval policy that selects informative observations based on their resulting long-term performance improvement, instead of similarity-based retrieval. To support continual improvement, EvoTune organizes tuning feedback into a hierarchical memory and incrementally refines both subspace localization and tuning policies without requiring LLM fine-tuning. Extensive experiments show that EvoTune consistently outperforms state-of-the-art baselines, achieving up to 44.5% performance improvement under the same tuning budget and reaching the best competing baseline's final performance up to 3.9X faster.

Problem

• DBMS tuning
• multi-component optimization
• combinatorial search space
• limited feedback
• suboptimal convergence

Innovation

• memory-aware evolution
• collaborative diagnosis
• utility-aware retrieval
• multi-component DBMS tuning
• hierarchical memory