Dynamic QoS optimization architecture for cloud-based DDDAS

Tools

Chen, T. ORCID: 0000-0001-5025-5472, Bahsoon, R. and Theodoropoulos, G., 2013. Dynamic QoS optimization architecture for cloud-based DDDAS. Procedia Computer Science, 18, pp. 1881-1890. ISSN 1877-0509

Preview

Text
PubSub10099_Chen.pdf - Post-print
Download (250kB) | Preview

Official URL: http://doi.org/10.1016/j.procs.2013.05.357

Abstract

Cloud computing urges the need for novel on-demand approaches, where the Quality of Service (QoS) requirements of cloud-based services can dynamically and adaptively evolve at runtime as Service Level Agreement (SLA) and environment changes. Given the unpredictable, dynamic and on-demand nature of the cloud, it would be unrealistic to assume that optimal QoS can be achieved at design time. As a result, there is an increasing need for dynamic and self- adaptive QoS optimization solutions to respond to dynamic changes in SLA and the environment. In this context, we posit that the challenge of self-adaptive QoS optimization encompasses two dynamics, which are related to QoS sensitivity and conflicting objectives at runtime. We propose novel design of a dynamic data-driven architecture for optimizing QoS influenced by those dynamics. The architecture leverages on DDDAS primitives by employing distributed simulations and symbiotic feedback loops, to dynamically adapt decision making metaheuristics, which optimizes for QoS tradeoffs in cloud-based systems. We use a scenario to exemplify and evaluate the approach.

Item Type:

Journal article

Alternative Title:

Dynamic Data-Driven Architecture for adaptive QoS optimization in the cloud

Description:

2013 International Conference on Computational Science

Publication Title:

Procedia Computer Science

Creators:

Chen, T., Bahsoon, R. and Theodoropoulos, G.

Publisher:

Elsevier