L2T-Tune:LLM-Guided Hybrid Database Tuning with LHS and TD3

TL;DR

L2T-Tune introduces a three-stage, LLM-guided hybrid framework for database configuration tuning that significantly improves performance, accelerates convergence, and enhances transferability across hardware and workload changes.

Contribution

It presents a novel three-stage tuning pipeline combining warm-start sampling, LLM-guided hint mining, and RL-based fine-tuning, addressing stability, convergence speed, and transferability issues.

Findings

Achieves 37.1% average performance improvement over state-of-the-art methods.

Reaches optimal tuning results in just 30 online steps.

Demonstrates rapid offline convergence comparable to RL models.

Abstract

Configuration tuning is critical for database performance. Although recent advancements in database tuning have shown promising results in throughput and latency improvement, challenges remain. First, the vast knob space makes direct optimization unstable and slow to converge. Second, reinforcement learning pipelines often lack effective warm-start guidance and require long offline training. Third, transferability is limited: when hardware or workloads change, existing models typically require substantial retraining to recover performance. To address these limitations, we propose L2T-Tune, a new LLM-guided hybrid database tuning framework that features a three-stage pipeline: Stage one performs a warm start that simultaneously generates uniform samples across the knob space and logs them into a shared pool; Stage two leverages a large language model to mine and prioritize tuning hints from manuals and community documents for rapid convergence. Stage three uses the warm-start sample pool to reduce the dimensionality of knobs and state features, then fine-tunes the configuration with the Twin Delayed Deep Deterministic Policy Gradient algorithm. We conduct experiments on L2T-Tune and the state-of-the-art models. Compared with the best-performing alternative, our approach improves performance by an average of 37.1% across all workloads, and by up to 73% on TPC-C. Compared with models trained with reinforcement learning, it achieves rapid convergence in the offline tuning stage on a single server. Moreover, during the online tuning stage, it only takes 30 steps to achieve best results.