Maximizing profit in cloud computing using M/G/c/k queuing model

Abstract

The economics of cloud computing is more and more a crucial issue. To maximize the profit, a service provider should understand both service charges and business costs, and how they are determined by the characteristics of the applications and the configuration of a multiserver system. In this work, we treat a multiserver system as an M/G/c/k queuing model with impatient customers, such that our optimization problem can be formulated and solved. There are no analytical formulas to calculate performance measures of M/G/c/k queue. Therefore, we provide new analytical formulas to compute the mean waiting time of new service request arrival to the system and the delay probability, i.e., the probability that a newly submitted service request must wait because all servers are busy. We focus on the calculation of these latter, because our performance metric in this work is the task waiting time. Then, expected revenue of a service provider is calculated. Numerical calculations of the optimal server size and the optimal server speed are demonstrated. Finally, the expected profit in one unit of time for a cloud provider is obtained.