Home // ICAS 2016, The Twelfth International Conference on Autonomic and Autonomous Systems // View article

QoS-Aware Scale up on IaaS Clouds

Authors:
Luis Fernando Orleans
Geraldo Zimbrão da Silva

Keywords: scale up, PROFUSE, IaaS cloud

Abstract:
For systems hosted in IaaS clouds that target profit, like blogs and e-commerce systems, the final revenue should be the most important metric. Balancing the QoS experienced by clients as a manner to avoid their drop-out against the cost related to leasing instances forms the basis of a good instance management for those scenarios. In this paper, we demonstrate the feasibility of using a Fuzzy Logic Inference System as a tool for maintaining the QoS. Furthermore, it was created a method for reducing the overall cost related to instances leasing using an incremental algorithm, acquiring one computational unit at a time. Finally, we used hooks to handle unpredictable burst of requests, called here as noises. Our experiments evidenced that those methods can keep the response time of all requests below a deadline, avoiding customer dissatisfaction. At the same time, the total cost of servers lease is reduced, even when the cost-benefit among different configurations is not linear.

Pages: 4 to 10

Copyright: Copyright (c) IARIA, 2016

Publication date: June 26, 2016

Published in: conference

ISSN: 2308-3913

ISBN: 978-1-61208-483-1

Location: Lisbon, Portugal

Dates: from June 26, 2016 to June 30, 2016