TY - JOUR
T1 - WWW Retrieval Handling Optimization wρ3
T2 - A metric for webpage timeout setting performance evaluation and comparison
AU - Seeling, Patrick
N1 - Publisher Copyright:
© 2019 Elsevier B.V.
PY - 2020/9
Y1 - 2020/9
N2 - The rise of HTTP-based communications and the employment of HTML5 for cross-platform application development gives rise to a continuous increase of traffic over all types of networks and a continuous demand for caching of content. However, a remaining underlying problem is the timeout of resources, typically configured server-side and dependent on the type of web site that offers content. With common web server implementations, individual webobjects constituting the overall content cannot be readily configured independently, which results in premature timeouts of webobjects with undesired consequences on download times, rendering speed, and energy needed. We present and evaluate WWW Retrieval Handling Optimization wρ3, which provides optimization potentials for webpages. wρ3 can also be used to derive hints for fine-tuning the performance of the server-side configurations and to compare the performance of timeouts between web sites of different categories. Our metric features a tunable preference between counts of webobjects and the amount of data required and is a function of the re-visiting time. We showcase the applicability of wρ3 over a range of different web sites and categories, additionally evaluating the applicability over time horizons and worldwide request locations. The overarching flexibility and ease of use together with its overall performance evaluation results make wρ3 an ideal candidate as a generic web site performance measurement tool.
AB - The rise of HTTP-based communications and the employment of HTML5 for cross-platform application development gives rise to a continuous increase of traffic over all types of networks and a continuous demand for caching of content. However, a remaining underlying problem is the timeout of resources, typically configured server-side and dependent on the type of web site that offers content. With common web server implementations, individual webobjects constituting the overall content cannot be readily configured independently, which results in premature timeouts of webobjects with undesired consequences on download times, rendering speed, and energy needed. We present and evaluate WWW Retrieval Handling Optimization wρ3, which provides optimization potentials for webpages. wρ3 can also be used to derive hints for fine-tuning the performance of the server-side configurations and to compare the performance of timeouts between web sites of different categories. Our metric features a tunable preference between counts of webobjects and the amount of data required and is a function of the re-visiting time. We showcase the applicability of wρ3 over a range of different web sites and categories, additionally evaluating the applicability over time horizons and worldwide request locations. The overarching flexibility and ease of use together with its overall performance evaluation results make wρ3 an ideal candidate as a generic web site performance measurement tool.
KW - Caching
KW - Information retrieval
KW - Information science
KW - Web performance
KW - Web sites
UR - http://www.scopus.com/inward/record.url?scp=85076222426&partnerID=8YFLogxK
U2 - 10.1016/j.future.2019.11.023
DO - 10.1016/j.future.2019.11.023
M3 - Article
AN - SCOPUS:85076222426
SN - 0167-739X
VL - 110
SP - 1055
EP - 1066
JO - Future Generation Computer Systems
JF - Future Generation Computer Systems
ER -