City Research Online

A min-max framework for CPU resource provisioning in virtualized servers using ℋ∞ Filters

Charalambous, T. and Kalyvianaki, E. (2010). A min-max framework for CPU resource provisioning in virtualized servers using ℋ∞ Filters. Decision and Control (CDC), 2010 49th IEEE Conference on, pp. 3778-3783. doi: 10.1109/CDC.2010.5717375

Abstract

Dynamic resource provisioning for virtualized server applications is integral to achieve efficient cloud and green computing. In server applications unpredicted workload changes occur frequently. Resource adaptation of the virtual hosts should dynamically scale to the updated demands (cloud computing) as well as co-locate applications to save on energy consumption (green computing). Most importantly, resource transitions during workload surges should occur while minimizing the expected loss due to mismatches of the resource predictions and actual workload demands. Our approach is to minimize the maximum expected loss using the same techniques as in two-person zero-sum games. We develop an H∞ filter that minimizes the worst-case estimation and allocate resources fast. Through simulations our H∞ filter demonstrates its effectiveness and good performance when compared against Kalman-based controllers.

Publication Type: Article
Additional Information: © 2010 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.
Subjects: Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Departments: School of Mathematics, Computer Science & Engineering > Computer Science
URI: http://openaccess.city.ac.uk/id/eprint/8178
[img]
Preview
Text - Accepted Version
Download (826kB) | Preview

Export

Downloads

Downloads per month over past year

View more statistics

Actions (login required)

Admin Login Admin Login