| 引用本文: | 吴志镛,梁杰,姚灵灵,庞枢,符景洲,张弛,李飞飞,姜宇.基于大语言模型的数据库管理系统可靠性测试技术研究.软件学报,2026,37(8):3257-3292 |
| |
|
| |
|
|
| 本文已被：浏览 380次 下载 170次 |
码上扫一扫！ |
|
|
| 基于大语言模型的数据库管理系统可靠性测试技术研究 |
|
吴志镛1, 梁杰1, 姚灵灵2, 庞枢2, 符景洲1, 张弛1, 李飞飞2, 姜宇1
|
|
1.清华大学 软件学院, 北京 100084;2.阿里云计算有限公司, 浙江 杭州 310030
|
|
| 摘要: |
| 数据库管理系统是支撑现代信息基础设施的核心软件, 其可靠性直接影响数据安全与业务连续性. 随着系统复杂度的不断提升, 数据库中的缺陷可能导致数据损坏、信息泄露及系统崩溃等严重后果. 近年来, 模糊测试作为一种高效的自动化缺陷检测技术, 已在数据库可靠性测试中广泛应用并取得显著成果. 然而, 传统模糊测试方法常依赖于简单的规则和模式生成测试用例, 难以生成具有更深层次语义理解的复杂场景, 因而在覆盖数据库管理系统复杂交互路径与触发深层缺陷方面仍存在不足. 与此同时, 大语言模型(large language model, LLM)的快速发展为数据库测试带来了新的机遇. 凭借其强大的语义理解、上下文推理与自我学习能力, LLM能够生成多样化、语义合理的SQL测试用例, 辅助测试结果验证与缺陷分析, 显著提升数据库管理系统可靠性测试的自动化与智能化水平, 挖掘数据库深层次缺陷. 系统梳理大语言模型在数据库可靠性测试中的研究进展, 分析基于大语言模型的测试框架在测试用例生成、结果校验、覆盖反馈与测试优化等方面的最新成果, 评估现有研究的有效性与局限, 并展望未来数据库管理系统可靠性测试的发展方向. |
| 关键词: 大语言模型 数据库管理系统 漏洞挖掘 测试用例生成 |
| DOI：10.13328/j.cnki.jos.007653 |
| 分类号:TP311 |
| 基金项目:国家自然科学基金(625B2100, 62302256, 62525207, 62021002, U2441238); 中央高校基本科研业务费专项资金(JKF-2025023600609); CCF-阿里云瑶池科研基金(2024007) |
|
| Research on LLM-based Reliability Testing for Database Management Systems |
|
WU Zhi-Yong1, LIANG Jie1, YAO Ling-Ling2, PANG Shu2, FU Jing-Zhou1, ZHANG Chi1, LI Fei-Fei2, JIANG Yu1
|
|
1.School of Software, Tsinghua University, Beijing 100084, China;2.Alibaba Cloud Computing Co. Ltd., Hangzhou 310030, China
|
| Abstract: |
| Database management systems (DBMSs) are the cornerstone of modern information infrastructure, and their reliability directly impacts data security and business continuity. As system complexity increases, bugs in DBMSs can lead to data corruption, information leakage, or even system failures. In recent years, fuzzing has become an efficient automated technique for detecting bugs and has been widely used in DBMS reliability testing, yielding significant results. However, traditional fuzzing approaches typically rely on simple rules or pattern-based test case generation, which makes it difficult to construct complex scenarios with deeper semantic understanding. Consequently, they remain insufficient for covering complex interaction paths and triggering deep-seated bugs in DBMSs. Meanwhile, the rapid advancement of large language models (LLMs) has brought new opportunities for DBMS testing. With strong semantic understanding, contextual reasoning, and self-learning capabilities, LLMs can generate diverse and semantically valid SQL test cases, assist with result validation and defect analysis, and significantly improve the automation and intelligence of DBMS reliability testing, enabling the discovery of deep-seated defects in databases. This paper presents a systematic review of research progress on the application of LLMs in DBMS reliability testing. The latest advances in LLM-based testing frameworks are analyzed in terms of test case generation, result validation, coverage feedback, and testing optimization. The effectiveness and limitations of existing studies are evaluated, and future development directions for DBMS reliability testing are discussed. |
| Key words: large language model (LLM) database management system (DBMS) vulnerability discovery test case generation |
|
|
|
|
