Kepler: Robust Learning for Parametric Query Optimization
Lyric Doshi, Vincent Zhuang, Gaurav Jain et al. · Proceedings of the ACM on Management of Data · 2023
Most existing parametric query optimization (PQO) techniques rely on traditional query optimizer cost models, which are often inaccurate and result in suboptimal query performance. We propose Kepler, an end-to-end learning-based approach to PQO that demonstrates significant speedups in query latency over a traditional query optimizer. Central to our method is Row Count Evolution (RCE), a novel plan generation algorithm based on perturbations in the sub-plan cardinality space. While previous approaches require accurate cost models, we bypass this requirement by evaluating candidate plans via actual execution data and training anML model to predict the fastest plan given parameter binding values. Our models leverage recent advances in neural network uncertainty in order to robustly predict faster plans while avoiding regressions in query performance. Experimentally, we show that Kepler achieves significant improvements in query runtime on multiple datasets on PostgreSQL.
Cited by 35 works. doi:10.1145/3588963 Open-access PDF
Authors
- Lyric Doshi
- Vincent Zhuang
- Gaurav Jain
- Ryan Marcus
- Haoyu Huang
- Deniz Altınbüken
- Eugene Brevdo
- Campbell B. Fraser
References
- Multi-Label Classification
- Parametric query optimization
- How good are query optimizers, really?
- OLTP-Bench
- Deep Reinforcement Learning for Join Order Enumeration
- Learning-based Query Performance Modeling and Prediction
- Bao: Making Learned Query Optimization Practical
- Example-Dependent Cost-Sensitive Logistic Regression for Credit Scoring
- Parametric Query Optimization for Linear and Piecewise Linear Cost Functions
- AI Meets AI