Model-Based Data Centre Costing Approach

Abstract

Data centres form a key part of the infrastructure upon which a variety of information technology services are built. They provide the capabilities of centralized repository for storage, management, dissemination and networking of data and information organized around a particular body of knowledge or pertaining to a particular business. The increasing demand for data storage and compute resources has driven the growth of large data centres: the massive server farms that run many of our present day internet and e-commerce applications with its attendant significant investment capital outlay and ongoing operating costs. This makes a critical examination of costs associated with building, operating and upgrading data centres imperative. For this cost model to meet the requirements of stakeholders it must capture the multidimensional complexities in power delivery, cooling, and required levels of redundancies for a specific service level agreement. However, non factoring of major total cost of ownership (TCO) parameters like amortization and maintenance costs for power, cooling and information technology resources in the cost model for data centres have questioned the reliability of this cost model, hence necessitate a research of this nature. To attack the problems identified above, I adopted a more accurate data centre construction cost model developed at HP Labs and adopted by the Uptime Institute Inc., to serve as a business intelligence tool for measurement and prediction of TCO and return-on-investment analysis.

Keywords: Data centres, cost model, service level agreement, TCO.