In the terminology of computer networks, the cloud is defined as a cluster of computing resources, such as software, storage, processor, memory etc. The cloud serves these resources as a service to the users (consumers) through a network (most commonly using internet). The hiring of cloud resources reduces the financial requirements necessary for the development and maintenance of infrastructure. The cloud is provided as a service, hence it must serve the users request quickly while efficiently

more »

... utilizing the resources to keep operational cost as low as possible. The load balancer is responsible for effectively performing above. In simple terms the job of the load balancer is to distribute the traffic coming from different users on the available resources in such an optimal way that maintains the QoS while minimizes the operational cost. While the job of load balancer is quite complex in this paper in we modeled it using queuing theory to evaluate the response time for services. The model can also be used to optimally scale the cloud system to guarantee the QoS for given response time, and efficient formation of VMs (virtual machines) based on the system load. We further evaluated the different scheduling models using numerical analysis and compare them with the queuing model derived. The rest of the paper is organized as follows. The second section presents an overview on the related topic. The third section describes the cloud computing while the modeling is presented in fourth section. The simulation results are presented in fifth section followed by the conclusion in the sixth section. Aim of the Study This paper is aimed to (1) analyze the potential of applicability of queuing theory in the cloud system for the purpose of job scheduling and efficient resource utilization. (2) Development of accurate and reliable numerical simulation model for the analysis of complex queuing model development and analysis. Review of Literature Vacation queuing theory based energy saving task scheduling in a heterogeneous cloud computing system is described by Chunling Cheng et al. 1 which deals in minimizing the power loss on the processing nodes remains powered on all the time to await incoming tasks, because the randomness in incoming jobs in cloud computing environments. Hamzeh Khazaei et al. 2 modeled a cloud system as an M/G/m queuing system, which means considering the inter-arrival time of requests is exponentially Abstract The cloud computing represents a concept of networking where the computing resources and users are located at different places, and the resources (such as hardware, softwareetc.)are accessible to the cloud users as a service through a computer network. To maintain efficient utilization of its resources while providing better QoS (which includes minimization of pending jobs in queue, faster job execution); the load balancer uses different algorithms for job scheduling and resource utilization. In this paper we analyzed the different job scheduling algorithms used by load balancer in terms of Queue Length, Waiting Time, and Load Distribution for Different Traffic Models using Matlab.