The rise in computer storage and network performance means that disk I/O and network communication are often no longer bottlenecks in database management systems (DBMSs). Instead, the overheads associated with operating system (OS) services (e.g., system calls, thread scheduling, and data movement from kernel-space) limit query processing responsiveness. To avoid these overheads, user-space applications prioritizing performance over simplicity can elide these software layers with a kernel-bypass design. However, extracting benefits from kernel-bypass frameworks is challenging, and the libraries are incompatible with standard deployment and debugging tools. For these reasons, few DBMSs employ a kernel-bypass approach.

We present user-bypass — an approach to designing DBMS software that complements OS extensibility. With user-bypass, developers write safe, event-driven programs to push DBMS logic into the kernel’s stack using eBPF and avoid user-space overheads.

People

• Matt Butrovich
• John Rollinson
• Justine Sherry
• Andy Pavlo

Acknowledgements

This project supported (in part) by Google, VMWare, Amazon, Alfred P. Sloan Research Fellowship, and the U.S. National Science Foundation (SPX-1822933).

Publications

1. M. Butrovich, K. Ramanathan, J. Rollinson, W. S. Lim, W. Zhang, J. Sherry, and A. Pavlo, "Tigger: A Database Proxy That Bounces With User-Bypass," Proc. VLDB Endow., vol. 16, iss. 11, pp. 3335-3348, 2023. PDF Bibtex

   @article{butrovich23,
      author = {Butrovich, Matthew and Ramanathan, Karthik and Rollinson, John and Lim, Wan Shen and Zhang, William and Sherry, Justine and Pavlo, Andrew},
      title = {Tigger: A Database Proxy That Bounces With User-Bypass},
      journal = {Proc. {VLDB} Endow.},
      volume = {16},
      number = {11},
      pages = {3335--3348},
      year = {2023},
      url = {https://www.vldb.org/pvldb/vol16/p3335-butrovich.pdf},
    }

2. M. Butrovich, W. S. Lim, L. Ma, J. Rollinson, W. Zhang, Y. Xia, and A. Pavlo, "Tastes Great! Less Filling! High Performance and Accurate Training Data Collection for Self-Driving Database Management Systems," in Proceedings of the 2022 International Conference on Management of Data, 2022, pp. 617-630. PDF Bibtex

   @inproceedings{butrovich22,
      author = {Butrovich, Matthew and Lim, Wan Shen and Ma, Lin and Rollinson, John and Zhang, William and Xia, Yu and Pavlo, Andrew},
      title = {Tastes Great! Less Filling! High Performance and Accurate Training Data Collection for Self-Driving Database Management Systems},
      year = {2022},
      booktitle = {Proceedings of the 2022 International Conference on Management of Data},
      pages = {617--630},
      numpages = {14},
      series = {SIGMOD '22},
      doi = {10.1145/3514221.3517845},
      url = {https://db.cs.cmu.edu/papers/2022/moddm074-butrovich.pdf},
    }