Blackbox prediction of the impact of DVFS on end-to-end performance of multitier systems

Dynamic voltage and frequency scaling (DVFS) is a wellknown technique for gaining energy savings on desktop and laptop computers. However, its use in server settings requires careful consideration of any potential impacts on endto-end service performance of hosted applications. In this paper, we develop a simple metric called the “frequency gradient” that allows prediction of the impact of changes in processor frequency on the end-to-end transaction response times of multitier applications. We show how frequency gradients can be measured on a running system in a pushbutton manner without any prior knowledge of application semantics, structure, or configuration settings. Using experimental results, we demonstrate that the frequency gradients provide accurate predictions, and enable end-to-end performance-aware DVFS for mulitier applications.

• Type: Journal
• Year: 2010
• Where: SIGMETRICS
• Authors: Shuyi Chen, Kaustubh R. Joshi, Matti A. Hiltunen, Richard D. Schlichting, William H. Sanders
• Keywords: End-to-end Transaction Response, Endto-end Service Performance, Gradients, Hardware, SIGMETRICS 2010