PrePrint: Mathematical Modeling of Performance and Utility Consumption for A Heterogeneous Cloud Computing Environment

Mathematical Modeling of Performance and Utility Consumption for A Heterogeneous Cloud Computing Environment

Cloud computing paradigm has emerged as a highly cost-effective computation model for IT enterprise applications, scientific computing, and personal data management. The components in a cloud infrastructure are heterogeneous by nature and continue to be renovated during system upgrades, providing a plethora of physical processing nodes with a wide range of capability, performance, power consumption, and thermal characteristics. As such, to perform a goal-oriented provisioning for workloads and to understand their trade-off in a cloud computing facility is a challenging task. In this paper, we analyze a parallelized task dispatched and executed on a heterogeneous cloud infrastructure and use an ideal mathematical model to show the trade-off using energy-delay product as the objective metric by taking the execution time and energy consumption into account simultaneously. As we show, to deliver the optimal performance per utility consumed in a heterogeneous cloud, the response time of the slowest node (b) divided by that of the fastest node (a) should be less than or equal to three (b/a=3). Our theoretical models and analysis results can be used to guide allocation, deployment, and upgrades of computing nodes to optimize utility effectiveness in cloud computing services.

Index Terms: B.9.2 Energy-aware systems, C.1.4 Parallel Architectures, C.2 Communication/Networking and Information Technology, Multimedia Information Systems, H.3.5 Online Information Services, web programming, web applications, Pervasive computing, mobile applications, Internet applications, database technology, Computer-supported cooperative work, B Hardware, C Computer Systems Organization, D Software/Software Engineering, I.4 Image Processing and Computer Vision, B.9.2 Energy-aware systems, C.0.a Emerging technologies, C.0.e System architectures, integration and modeling,Citation: Sungkap Yeo, Hsien-Hsin Lee, "Mathematical Modeling of Performance and Utility Consumption for A Heterogeneous Cloud Computing Environment," Computer, 25 Mar. 2011. IEEE computer Society Digital Library. IEEE Computer Society, Usage of this product signifies your acceptance of the Terms of Use.

View the original article here

This post was made using the Auto Blogging Software from WebMagnates.org This line will not appear when posts are made after activating the software to full version.