Determining the Accuracy Required in Resource Load Prediction to Successfully Support Application Agility
Loading...
Authors
Hensgen, Debra
Kidd, Taylor
Xie, Geoffrey
Kresho, John
Subjects
Advisors
Date of Issue
2013-08-13
Date
Publisher
Language
Abstract
Largely due to the proliferation of the World Wide Web, and interfaces such as Netscape, users expect to have many different types of information immediately available. When they encounter a lengthy delay caused byheavy loads on shared resources, such as networks or servers, users often (manually) adapt by requesting different forms of the same information. As both mobile and agent computing becomes more popular, users will expect their applications to automatically adapt to heavy resource loads by fetching the information in a different form, e.g., text instead of graphics. This paper studies the accuracy with which resource loading information, particularly network loading information, must be known in order for applications to successfully, and with agility, adapt. We determine that under many normal conditions, fairly inaccurate estimates of currently available bandwidth suffice. However, when the system is heavily loaded, some strategies can perform much better with very accurate load estimates. That is, assuming that the adaptive applications have hard deadlines for obtaining the data they request, up to more of them will receive some form of that data on time, if the adaptation strategy has a good estimate of avail able bandwidth . Additionally, in these situations, ap plications that have a better estimate of bandwidth can deliver, on average, larger sized messages corres ponding to, in many cases, higher quality data. Fi nally, the accuracy with which the bandwidth must be known varies not only with inter arrival rate, but also with the adaptation strategy used and the percentage of adaptive applications in the system .
Type
Description
Series/Report No
Department
Identifiers
NPS Report Number
Sponsors
Funder
Format
Citation
Distribution Statement
Rights
This publication is a work of the U.S. Government as defined in Title 17, United States Code, Section 101. Copyright protection is not available for this work in the United States.