A Model for Cost-Benefit Analysis of Cloud Computing

INTRODUCTION Cloud computing paradigm has evolved recently and it has taken commercial computing to a new level. The concept of cloud computing rests upon the idea that computing resources will reside somewhere other than the computer room and that the users will connect to it using the resources as and when required. In effect, it displaces the infrastructure to the network so that the overall cost with respect to the management of hardware/software resources is reduced (Hayes, 2008). It appears to be highly disruptive technology (Rimal et al., 2009) hinting to the future where computation moves from local computers to centralized facilities operated by third party compute and storage utilities (Foster et al., 2008). However, considering the practical implementation of cloud computing, the adoption of cloud platforms by organizations/scientific community is in its infancy. There is a paucity of research towards a model that can demonstrate the benefits of cloud computing adoption and suggest the ideal time to shift to cloud computing. This study attempts to develop a cost-benefit analysis model that can present a clear picture to the IT managers when the shifting from legacy systems to cloud computing is concerned. The computing resources and IT infrastructure of every organization is idiosyncratic. Hence, direct recommendations on profitability cannot be given until all the inputs of organizations have been considered for profitability evaluation. This paper, therefore, suggests a model that can take various parameters of an organization and provide recommendations on profitability of shifting to cloud computing. The pricing model in cloud computing is quite similar to usage based pricing. Customers pay for the computing resources by means of customized service level agreement hiding the underlying technological infrastructure (Xiong & Perros, 2009). This concept of pay-as-you-go in cloud computing differs from traditional renting method which involves payment of negotiated cost to have the resource for a specific period of time irrespective of the actual usage. The cloud computing service taken up in this paper for the cost-benefit analysis model is Amazon AWS. It rounds up their billing to the nearest server hour/GB per month. AWS is chosen as an example because major players like Amazon reflect the most common pricing mechanism in the cloud market. The proposed model in this paper works on three layers. These layers represent the different levels at which organizations plan to adopt cloud computing. In the first case, a base cost estimation is done where the organization can compare the cost of the entire computational facility in-house to the cost of shifting completely to loud computing. The second layer performs the analysis based on the data pattern expressed in terms of the average amount of data it processes, transfer rate, the demand estimation/provision etc. This layer gives an instant recommendation on the feasibility of shifting to the cloud by taking into account the inputs of layer one as well. The last layer is a project specific layer which would be very helpful for organizations planning to keep the present infrastructure intact and using cloud computing for a specific upcoming project. This third layer demonstrates the widely used scenario in present day organizations. It takes inputs concerning the nature of the project and give recommendations on executing the project on cloud. This paper proceeds as follows. Next section attempts derive acumen from past studies that have discussed cost benefit analysis of cloud computing. This is followed by three layers of the model. Layer 1 describes the variables and the methods used for the computation of base cost estimation. Layer 2 attempts the same for data pattern variables followed by Layer 3 describing the project specific variables. The next section describes the proposed model using a three layer approach and how it can be used by organizations to get recommendations on cloud computing adoption. …