Ultron: History-Based Query Optimization at Databricks
Summary: Ultron brings history-based, subplan-level query optimization to Databricks’ massive lakehouse, compensating for sparse statistics by learning from repetitive workloads. Its drift-aware, three-tier architecture safely complements AQE, improving eligible join latency 25% in production. (summarized by gpt-5.6-luna on Aug 28 2026)
Incoming Non-self Citations Over Time
No non-self incoming citations found for this paper in this database.
Authors
- 1. Supun Nakandala (Databricks)
- 2. RK Korlapati (Databricks)
- 3. Eric Liang (Databricks)
- 4. Yunjia Zhang (Databricks)
- 5. Yuhao Zhang (Databricks)
- 6. Andrey Gubichev (Databricks)
- 7. Kelvin Jiang (Databricks)
- 8. Yannis Sismanis (Databricks)
- 9. Mostafa Mokhtar (Databricks)
- 10. Sid Taneja (Databricks)
- 11. Vuk Ercegovac (Databricks)
- 12. Amir Hormati (Databricks)
- 13. Amit Shukla (Databricks)
- 14. Michalis Petropoulos (Databricks)
- 15. Reynold Xin (Databricks)
- 16. Ryan Marcus (University of Pennsylvania)
BibTeX Citation
@article{nakandala_vldb26,
title = {{Ultron: History-Based Query Optimization at Databricks}},
author = {Nakandala, Supun and Korlapati, RK and Liang, Eric and Zhang, Yunjia and Zhang, Yuhao and Gubichev, Andrey and Jiang, Kelvin and Sismanis, Yannis and Mokhtar, Mostafa and Taneja, Sid and Ercegovac, Vuk and Hormati, Amir and Shukla, Amit and Petropoulos, Michalis and Xin, Reynold and Marcus, Ryan},
journal = {PVLDB},
series = {{VLDB} '26},
volume = {19},
number = {12},
pages = {4358--4371},
doi = {10.14778/3827998.3828038},
url = {https://doi.org/10.14778/3827998.3828038},
year = {2026}
}Incoming Citations (Sorted by Pagerank)
Showing 0 of 0 citing papers.
| Rank | Citing Paper | Year | Venue | Pagerank |
|---|
Previous
Page 1 / 1
Next
Outgoing Citations (Sorted by Pagerank)
Showing 1 of 51 cited papers.
Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.
| Rank | Cited Paper | Year | Venue | Pagerank |
|---|---|---|---|---|
| 10,596 | SEFRQO: A Self-Evolving Fine-Tuned RAG-Based Query Optimizer | 2026 | SIGMOD | 4.9793485e-05 |
Semantically Similar Papers
| # | Overall Rank | Paper | Year | Venue |
|---|---|---|---|---|
| 1 | 10,672 | PTO: A Workload-driven Predictive Table Optimizer for Lakehouse Systems | 2026 | SIGMOD |
| 2 | 2,662 | End-to-end Optimization of Machine Learning Prediction Queries | 2022 | SIGMOD |
| 3 | 5,039 | Leveraging Query Logs and Machine Learning for Parametric Query Optimization | 2022 | VLDB |
| 4 | 5,551 | Aero: Adaptive Query Processing of ML Queries | 2025 | SIGMOD |
| 5 | 4,069 | Dynamically Optimizing Queries over Large Scale Data Platforms | 2014 | SIGMOD |
| 6 | 10,916 | AutoLiquid: Autonomic Data Layout Optimization for the Databricks Lakehouse | 2026 | VLDB |
| 7 | 1,487 | Photon: A Fast Query Engine for Lakehouse Systems | 2022 | SIGMOD |
| 8 | 8,004 | A Spark Optimizer for Adaptive, Fine-Grained Parameter Tuning | 2024 | VLDB |
| 9 | 9,113 | Presto’s History-based Query Optimizer | 2024 | VLDB |
| 10 | 5,348 | Adaptive and Robust Query Execution for Lakehouses at Scale | 2024 | VLDB |