CATune: Structural Constraint-Aware Bayesian Optimization for DBMS Configuration Tuning

2026-10-06

AI Summary

This study addresses the inefficiency and susceptibility to infeasible regions in high-dimensional DBMS configuration tuning, where existing methods overlook deterministic dependency constraints. To this end, we propose CATune, a constraint-aware Bayesian optimization framework. CATune leverages large language models to accurately extract ordinal constraints from documentation, explicitly modeling them as topological structures within the search space. A topology-aware sampling strategy is further designed to perform direct optimization within feasible subspaces, achieving a paradigm shift from implicit learning to explicit structured constraint modeling. Experimental results demonstrate that CATune accelerates convergence by 12.5× over baselines and improves throughput by up to 63.37%, significantly enhancing system stability and tuning robustness.

Abstract

Modern DBMSs expose hundreds of configuration knobs, resulting in a high-dimensional and heterogeneous search space that makes automated tuning costly. Existing ML-based tuning systems typically treat the configuration domain as box-constrained and rely on workload feedback to implicitly capture inter-knob relationships. However, DBMS documentation specifies deterministic knob dependency constraints, particularly ordering constraints, that characterize structurally valid regions of the configuration space. We present CATune, a constraint-aware Bayesian optimization (BO) framework that models deterministic inter-knob ordering constraints as structural components of the search domain. Instead of learning feasibility boundaries through sampled violations, CATune performs optimization within a constraint-consistent subspace. We develop a topology-aware sampling strategy that respects dependency structure during exploration and avoids the inefficiencies of post-hoc constraint handling. To enable automated constraint discovery, we further design a precision-first extraction pipeline that combines LLM-based parsing with reliability safeguards to mitigate hallucinated dependencies. Experiments on PostgreSQL and MySQL using TPC-C and TPC-H workloads show that CATune substantially improves both sample efficiency and final tuning quality across surrogate models and BO frameworks. Under default ranges, CATune reaches the baseline optimum up to 12.5x faster and improves throughput by up to 63.37%. The improvements persist under knowledge-guided reduced ranges and alternative optimization implementations. These results demonstrate that explicitly modeling system-defined deterministic ordering constraints enhances optimization robustness and system stability.

Problem

• DBMS configuration tuning
• Bayesian optimization
• structural constraints
• knob dependency
• high-dimensional search space

Innovation

• Bayesian Optimization
• Constraint-Aware Tuning
• Topology-Aware Sampling
• LLM-based Extraction
• DBMS Configuration