From Presto to Prestissimo: A Velox-Powered Modernization Journey
Summary: Prestissimo replaces Presto’s Java workers with Velox-powered C++ workers, enabling an interruption-free, two-year migration at Meta. Production shadowing, fuzzing, and verification ensured feature parity, yielding 2–5× lower latency, less hardware, and 10× fewer reliability issues. (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. Amit Dutta (Meta)
- 2. Pedro Pedreira (Meta)
- 3. Xiaoxuan Meng (Meta)
- 4. Masha Basmanova (Meta)
- 5. Orri Erling (Meta)
- 6. Sergey Pershin (Meta)
- 7. Jialiang Tan (Meta)
- 8. Ge Gao (Meta)
- 9. Ke Wang (Meta)
- 10. Zac Wen (Meta)
- 11. Krishna Pai (Meta)
- 12. Xiao Du (Meta)
- 13. Natasha Sehgal (Meta)
- 14. Heidi Han (Meta)
- 15. Artem Selishchev (Meta)
- 16. Ke Wang (Meta)
- 17. Sebastiano Peluso (Meta)
- 18. Chandrashekhar Singh (Meta)
- 19. Sreeni Viswanandha (Meta)
- 20. Aditi Pandit (IBM; Meta)
- 21. Deepak Majeti (IBM; Meta)
- 22. Tim Meehan (IBM; Meta)
- 23. Christian Zentgraf (IBM; Meta)
- 24. Aakash Deep (Meta)
- 25. Rohit Jain (Meta)
- 26. Sridhar Anumandla (Meta)
- 27. Kaushik Ravichandran (Meta)
- 28. Vivek Gaur (Meta)
BibTeX Citation
@article{dutta_vldb26,
title = {{From Presto to Prestissimo: A Velox-Powered Modernization Journey}},
author = {Dutta, Amit and Pedreira, Pedro and Meng, Xiaoxuan and Basmanova, Masha and Erling, Orri and Pershin, Sergey and Tan, Jialiang and Gao, Ge and Wang, Ke and Wen, Zac and Pai, Krishna and Du, Xiao and Sehgal, Natasha and Han, Heidi and Selishchev, Artem and Wang, Ke and Peluso, Sebastiano and Singh, Chandrashekhar and Viswanandha, Sreeni and Pandit, Aditi and Majeti, Deepak and Meehan, Tim and Zentgraf, Christian and Deep, Aakash and Jain, Rohit and Anumandla, Sridhar and Ravichandran, Kaushik and Gaur, Vivek},
journal = {PVLDB},
series = {{VLDB} '26},
volume = {19},
number = {12},
pages = {4399--4412},
doi = {10.14778/3827998.3828041},
url = {https://doi.org/10.14778/3827998.3828041},
year = {2026}
}
Incoming Citations (Sorted by Pagerank)
Showing 1 of 1 citing papers.
| Rank | Citing Paper | Year | Venue | Pagerank |
|---|---|---|---|---|
| 10,929 | SQL-Native Vector Search at Billion Scale in Presto | 2026 | VLDB | 4.9793485e-05 |
Previous
Page 1 / 1
Next
Outgoing Citations (Sorted by Pagerank)
Showing 16 of 16 cited papers.
Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.
Previous
Page 1 / 1
Next
Semantically Similar Papers
| # | Overall Rank | Paper | Year | Venue |
|---|---|---|---|---|
| 1 | 4,302 | The Missing Piece in Complex Analytics: Low Latency, Scalable Model Management and Serving with Velox | 2015 | CIDR |
| 2 | 12,487 | Pronto: A Software-Defined Networking based System for Performance Management of Analytical Queries on Distributed Data Stores | 2014 | VLDB |
| 3 | 8,269 | Accelerate Distributed Joins with Predicate Transfer | 2025 | SIGMOD |
| 4 | 10,306 | Simple (yet Efficient) Function Authoring for Vectorized Engines | 2024 | VLDB |
| 5 | 7,451 | Modularis: Modular Relational Analytics over Heterogeneous Distributed Platforms | 2021 | VLDB |
| 6 | 6,670 | Composable Data Management: An Execution Overview | 2024 | VLDB |
| 7 | 10,929 | SQL-Native Vector Search at Billion Scale in Presto | 2026 | VLDB |
| 8 | 9,113 | Presto’s History-based Query Optimizer | 2024 | VLDB |
| 9 | 1,424 | Velox: Meta's Unified Execution Engine | 2022 | VLDB |
| 10 | 3,835 | Presto: A Decade of SQL Analytics at Meta | 2023 | SIGMOD |