Covering peak demand by using cloud services – an economic analysis

An optimization approach to capacity evaluation and investment decision of hybrid cloud: a corporate customer’s perspective

“…The optimal decision is based on the cost structure and a probability density function for computing demand in a given period. This model reflects customers' perspectives and utilizes a model presented by Henneberger [37] which extends the classic newsvendor problem and derives a critical fractile formula using an inverse cumulative distribution function of computing demand. This newsvendor problem has been widely used in inventory management.…”

“…x: an actual computing demand occurring in each time unit with an exponential probability distribution function λe −λx : an exponential probability distribution function for computing demand 1-e −λx : a cumulative exponential distribution function for computing demand c: units of private cloud capacity purchased at the beginning of a decision period k: one-time purchase cost per unit of the private cloud capacity for a decision horizon p: the price of the public cloud per time unit q: the guaranteed service level k a : the default downtime loss/penalty cost t: the number of time units in a decision horizon This section presents a base model for the cloud evaluation and investment decision problem. The following several assumptions used in this paper are from Henneberger [37]. It is assumed that private cloud resources are purchased or contracted at the beginning of the decision horizon, that the public cloud provider has a sufficient capacity to satisfy the peak demand of the company, and that the probability distribution for computing demand can be estimated based on real demand data and each demand can be split into the private and public clouds if cloud bursting is necessary.…”

“…where k ⋅ c is the total private cloud costs at the beginning of the investment horizon, q Á R ∞ c λe −λx ðx−cÞdx Á p is the costs of the public cloud per time unit, and ð1−qÞ Á R ∞ c λe −λx ðx−cÞdx Á k a is the default downtime penalty cost related to the out-of-service situation from the public cloud provider. The cost structure follows Henneberger [37]. Applying integration techniques, Eq.…”

Cybercrime and cloud computing. A game theoretic network model

“…In business‐to‐business relations, cloud computing becomes also more relevant as several advantages are realized (Armbrust et al, ; Boss, Malladi, Quan, Legregni, & Hall, ; Henneberger, ; Nazir, ): Costs are reduced because only actual usage has to be paid and no further software is needed, which in turn reduces the requirements for provision of powerful computers; transaction costs like coordination and information costs are reduced as data and software is available almost anytime and anywhere. However, there are disadvantages as well—especially with respect to security issues: Data are not stored in a closed local area (within the company's area of influence) but on (sometimes untrusted) servers somewhere else that have to be accessed via the Internet (which itself increases security issues; Abadi, ; Bisong & Rahman, ; Hashizume, Rosado, Fernández‐Medina & Fernandez, ); and of course, there is the danger of hold‐up as one has to rely on a third party (Bamiah & Brohi, ).…”