Performance Evaluation in Cloud Computing Model using Queuing Models

Abstract

Cloud computing is the process of enabling the network access to a set of selected users with good configuration of computing resources on the basis of availability of the network access whenever there is an demand for the services from the cloud. Cloud computing is a regular term and the regular service that was being delivering the required services to the hosts in the internet. Here, the cloud computing mechanism is used for describing both the list of platforms that were available to the users for working and also the several types of applications that can be processed. The present technique was being considered by most of the researchers and the research institutes as the most potential and the most useful area for the research and also for research in academia like universities and major research laboratories. Performance evaluation of several applications the related applications and their sub parts were being considered as one of the useful and mostly used research area in the recent years. This technique and its services were being used mostly for the providers of the cloud and its related areas and the beneficiaries of this technique of area were both the providers of the cloud and the customers related to the cloud. Only few notable works have been published with regards to performance evaluation in cloud computing. In general the analytical models were aimed at designing the models which use the cloud and its services through which the performance of the model was analyzed and evaluated under various configurations and assumptions. These assumptions were based on the queuing theory and its accuracy is verified with numerical calculations and simulations. Present paper deals with the performance evaluation in-terms of steady state parameters of a small cloud server farm using single and multi server queuing models. Single server models include M/M/1, M/G/1, M/D/1 and M/Er/1. Multi-server model considered include M/M/c, M/M/c/c, M/M/c/K and M/M/c/c+r. A comparison among the steady state parameters evaluated for the above queuing models with respect to traffic intensity is also presented.