Detecting Transactional Bugs in Database Engines via Graph-Based Oracle Construction

on YouTube ↗11m

" summary

A tool called TxCheck builds statement-dependency graphs via SQL-level instrumentation and uses transactional oracle construction to generate semantically equivalent, reordered test cases; the approach is proven correct for producing equivalent executions and catches divergence as a correctness failure rather than a crash. It runs against TiDB, MySQL and MariaDB, finds 56 unique bugs (52 confirmed, 18 fixed), and surfaces concrete failures such as a MySQL update-with-subselect misbehavior.

AI-written summary · about

" chapters (6)

1. 0:00Introduction to transactions
2. 2:16Challenges in database testing
3. 3:41Semantic preserving reordering
4. 5:58Methodology overview
5. 7:31Dependency graph analysis
6. 8:30Results and conclusion

" related talks

• 14mChecking Observational Correctness of Database SystemsAnkush Desai, Amanda Xu, Aws Albarghouthi, Lauren Pick, Sanjit A. Seshia · OOPSLA 2025
• 14mDetecting Logic Bugs in Database Engines via Equivalent Expression TransformationZhendong Su, Zu-Ming Jiang · 18th USENIX Symposium on Operating Systems Design and Implementation (OSDI '24)
• 10mHow to test software without writing testsManuel Rigger · It Will Never Work in Theory (Online, April 2022)
• 36mSQLancer: Automatically Finding Bugs in Database Management SystemsManuel Rigger · SPLASH 2020 REBASE
• 15mFinding Bugs in Database Systems via Query PartitioningManuel Rigger, Zhendong Su · OOPSLA 2020
• 55mMuch ado about testingNicolas Rinaudo · ScalaIO 2019
• 15mDDLumos: Understanding and Detecting Atomic DDL Bugs in DBMSsYu Jiang, Jie Liang, Zhiyong Wu, Jingzhou Fu, Wenqian Deng · USENIX ATC '25
• 20mGraphiti: Bridging Graph and Relational Database QueriesIşıl Dillig, Ruijie Fang, Yang He, Yuepeng Wang · PLDI 2025