Effective System Call Aggregation (ESCA)

The main objective of this work was to reduce the per-syscall overhead through the use of effective syscall aggregation. For that purpose, ESCA takes advantages of system call batching and exploits the parallelism of event-driven applications by leveraging Linux I/O model to overcome the disadvantages of previous solutions.

ESCA is capable of reducing the per-syscall overhead by up to 62% for embedded web servers. Real-world highly concurrent event-driven applications such as Nginx and Redis are known to benefit from ESCA, along with full compatibility with Linux syscall semantics and functionalities.

Demo

Nginx-ESCA led by about 11% over vanilla Nginx.

lighttpd-ESCA led by about 13% over vanilla lighttpd.

Presentation

We also present our work with topic: Reduce System Call Overhead for Event-Driven Servers at Open Source Summit Japan 2021, organized by the Linux Foundation.

• Presentation information
• slides
• video recording

Academic Paper

Please see our PDP 2022 paper, available in the IEEE Xplore digital library, and you can get a preprint copy.

Citation

If you find this work useful in your research, please cite:

@inproceedings{cheng2022esca,
    author={Cheng, Yu-Cheng and Huang, Ching-Chun (Jim) and Tu, Chia-Heng},
    booktitle={2022 30th Euromicro International Conference on Parallel, Distributed and Network-based Processing (PDP)},
    title={ESCA: Effective System Call Aggregation for Event-Driven Servers},
    year={2022},
    pages={18-25},
    doi={10.1109/PDP55904.2022.00012}
}