Search arXivSearch

arXiv · 2607.17841

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

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.

Explore related subjects

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Zhaoyan Hong, Yishen Sun, Xinyi Zhang, Zhentao Han, Jinhao Dong, Wei Lu, Kai Xu, Liu Tang, Qi Liu, Xiaoyong Du. 2026-07-20. From Blind Search to Memory-Aware Evolution: Efficient DBMS Tuning via Collaborative Diagnosis and Utility-Aware Retrieval. https://arxiv.org/abs/2607.17841

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

OmniTable: A Unified Wide-Table System for Petabyte-Scale LLM Data Curation and Exploration

Data curation is a critical bottleneck in industrial-grade LLM development, where petabyte-scale unstructured corpora are scattered across hundreds of physical tables, feature engineering relies on manual, table-centric pipeline orchestration, and data lineage is largely absent. We present OmniTable as an architecture blueprint for a unified wide-table layer built on Logical Unification, Physical Separation, targeting petabyte-scale LLM data curation and exploration. OmniTable makes four contributions: (1) a unified wide-table abstraction that consolidates multi-source heterogeneous data and thousands of derived features under a single logical schema via logical-physical mapping; (2) declarative feature lifecycle management that automates dependency resolution, execution planning, operator fusion, and lineage tracking, replacing manual pipeline orchestration with a "declare-and-execute" paradigm; (3) an adaptive execution engine with autonomous governance that achieves stable PB-scale feature backfill through heterogeneous compute routing (CPU/GPU), adaptive tuning, UDF-level fault tolerance, and automated storage layout optimization; and (4) hybrid-accelerated data exploration combining a global ID index, transparent OLAP offloading, and background materialized views to deliver second-level point lookups and filtered exports exceeding 20 TB/hour. In production, OmniTable manages over 35 PB of training data across web, code, PDF, and SFT domains, reducing the human-in-the-loop curation cycle from approximately 14 days to approximately 2.5 days (5.6x over the pre-OmniTable production workflow), with consistent feature versioning, auditable lineage, and minimal manual intervention.

cs.DB

A Theoretical Study of DBLog: Certified Virtual Cuts for a Snapshot-Equivalent Replay of Live Databases

DBLog is a change-data-capture (CDC) mechanism for copying a table or selected keys from a source database while continuing to stream new changes from its commit log. It reads a table in primary-key ordered chunks and brackets each read with low and high watermarks in the source log. A chunk row is discarded if a log event for the same key appears anywhere in that window. The remaining rows are emitted as refresh events after the high watermark. DBLog allows backfills to run at any time during normal operation, rather than only as an initial bootstrap, while writes and live capture continue. It was introduced in the 2019 Netflix Tech Blog post and further discussed in the 2020 DBLog paper. The mechanism has since been adopted by open-source projects, including Debezium and Apache Flink CDC. The original blog post and paper explained this mechanism but did not provide a formal correctness proof or a precise description of its replay result. In this paper, we formalize how chunk reads and log events are combined and prove that their replay reconstructs the source state up to a specific log position for the keys being copied. We define this as a virtual cut, representing a snapshot-equivalent replay without requiring a physical snapshot. DBLog emits the events needed to reconstruct the source state downstream, without storing that state itself. To verify executions independently, we provide a certificate that validates this equivalence from recorded log evidence and chunk observations. Once chunk processing completes and covers all keys, this guarantee extends to the whole table. Furthermore, appending subsequent change-stream events advances the cut forward, formalizing the seamless transition from historical backfill to live streaming. All definitions and proofs are mechanized in Isabelle/HOL.

cs.DB

When Does Low-Bit Quantization Preserve the Decisions of Vector Search?

Low-bit quantization can achieve high recall on some vector representations and fail sharply on others, while average distortion and global rank correlation do not explain the difference. We study quantized vector search at the level of the comparisons consumed by ranking and graph-pruning algorithms. Our first result is a distribution-free decomposition: the probability that a comparison flips is bounded by the probability mass of exact margins near zero plus the tail probability of the calibrated residual. We then account for dependence between residuals that share a query or graph node, and derive covariance-aware second-moment identities and tail bounds under a joint MGF proxy. For a frozen candidate permutation, we prove a deterministic coupling theorem for Vamana neighbour selection: the approximate replay returns the exact neighbour list exactly when all candidate-level pruning actions agree on the frozen exact states. We connect these results to representation geometry through an exact Gaussian oracle, establish a strict correlation gain from a deterministic magnitude bit in an aligned bilinear model, and give a rare-contamination construction showing why marginal Gaussian diagnostics do not imply the required residual tails. When analytical assumptions are unavailable, a held-out block certificate bounds the selective failure risk of a frozen quantized rule. Across learned, classical, and synthetic embeddings, standardized exact margins predict held-out ranking and pruning flip rates substantially better than global rank correlation. The framework applies to coordinate binary codes, RaBitQ, Lucene BBQ, and product quantizers through a common decision interface.

cs.DB