Abstract:
Since the high demand of web requests often stresses or even saturates systems that have very large capabilities and since there are very few published results addressing the issue of quantitatively characterizing the behavior of Web servers and improving their performance, the web server has become the bottle neck of the web performance. It’s worth noting that some of this saturation in web servers is due to “flash crowds”, in other words temporary, therefore having a cluster of web servers can be expensive yet unnecessary. This project aims at focusing on the lack of optimization in web servers. Since there are many performance issues with the internets leading web server, Apache, we decided to implement a new web server, based on the object oriented web server previously implemented last year, Artemis. By implementing a new design architecture and addressing many of the optimization issues that we researched, we hope to come up with a new prototype for a web server that will not only match Apache in performance, but outperform it to a reasonable extent.Thesis Statement:
“Develop a web server to outperform Apache 2.0.48 in terms of decreasing the request response time through handling a larger number of users and/or serving more requests per second.”Documents:
Presentations:
Graphs, Tables and Results:
ABYSS Compiled Object for Redhat/FedoraCore Linux Flavors:
Miscellaneous Information: