The Effects of HTTP Advances on Modern Page Loading

Author:

Pendon Jon1ORCID,Warishe Tejas1

Affiliation:

1. New Jersey Institute of Technology

Abstract

Since the dawn of the internet, protocols have seen massive innovations that have improved our experiences. The most prominent of these advances are in the application layer space with HTTP. HTTP has moved from HTTP 1.1 to HTTP 2 and now HTTP 3. The second most prominent occurred one layer higher in the transport layer. The main one being the shift from loss-based congestion control to delay-based congestion control. This paper argues that due to these changes in TCP settings, sometimes, the improvements from HTTP 1.1 to 2 or even 3 have seen lower web performance. Utilizing automated web drivers to load web pages, this paper finds the amount of time it takes to load web pages and quantifies it as a web performance metric. A script is created to crawl through the websites provided and records the amount of time it takes to load the web page’s resources. This paper analyzes 50 websites to start but is able to scale to as many as needed. From these 50 websites, 9 websites perform worse using HTTPS 1, 2 and 3 as the main protocol and 14 websites perform worse using HTTP 1, 2, and 3. This information contradicts some background papers that show a large portion of websites would perform better using HTTP 1 only. However, these papers were concluded many years ago. Since then, web servers have upgraded their servers and protocols have been optimized. In conclusion, this information is useful in understanding how individual web sites interact with the network protocols and can help aid in protocol development.

Publisher

ScienceOpen

Reference7 articles.

1. Mind the delay;Agarwal;Proceedings of the 16th International Conference on emerging Networking EXperiments and Technologies,2020

2. BBR: Congestion-Based Congestion Control;Cardwell;Queue,2016

3. Taming the elephants: New TCP slow start;Ha;Computer Networks,2011

4. Hypertext Transfer Protocol Version 2 (HTTP/2);M. Belshe,2015

5. TCP Fluid Modeling with a Variable Capacity Bottleneck Link;A. Baiocchi;IEEE INFOCOM 2007 - 26th IEEE International Conference on Computer Communications,2007

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3