Affiliation:
1. Carnegie Mellon University, Pittsburgh, PA
Abstract
This article provides a detailed implementation study on the behavior of web serves that serve static requests where the load fluctuates over time (transient overload). Various external factors are considered, including WAN delays and losses and different client behavior models. We find that performance can be dramatically improved via a kernel-level modification to the web server to change the scheduling policy at the server from the standard FAIR (processor-sharing) scheduling to SRPT (shortest-remaining-processing-time) scheduling. We find that SRPT scheduling induces no penalties. In particular, throughput is not sacrificed and requests for long files experience only negligibly higher response times under SRPT than they did under the original FAIR scheduling.
Publisher
Association for Computing Machinery (ACM)
Subject
Computer Networks and Communications
Reference64 articles.
1. Web content adaptation to improve server overload behavior
2. Almesberger W. Linux network traffic control---implementation overview. Available at http://lrcwww.epfl.ch/linux-diffserv/.]] Almesberger W. Linux network traffic control---implementation overview. Available at http://lrcwww.epfl.ch/linux-diffserv/.]]
3. Workload characterization of the 1998 world cup Web site;Arlitt M.;IEEE Network,2000
Cited by
55 articles.
订阅此论文施引文献
订阅此论文施引文献，注册后可以免费订阅5篇论文的施引文献，订阅后可以查看论文全部施引文献
1. Bouncer: Admission Control with Response Time Objectives for Low-latency Online Data Systems;Companion of the 2024 International Conference on Management of Data;2024-06-09
2. SFS: Smart OS Scheduling for Serverless Functions;SC22: International Conference for High Performance Computing, Networking, Storage and Analysis;2022-11
3. A reliability benchmark for actor-based server languages;Proceedings of the 21st ACM SIGPLAN International Workshop on Erlang;2022-09-06
4. Scheduling Challenges for Variable Capacity Resources;Job Scheduling Strategies for Parallel Processing;2021
5. Resampling with Feedback: A New Paradigm of Using Workload Data for Performance Evaluation;Job Scheduling Strategies for Parallel Processing;2021