Effective Utilization of Resources and throughput computation is an important consideration in using cloud resources. Each resource is allocated by the tasks and each resource has virtual machines to handle the subtasks of use cases. The cost performance analysis is done using the markov model and the Server Allocation Matrix (SAM). Resource allocation in cloud computing System is a new dimension in place by taking into consideration the failure of the physical machines in the datacenter that constitutes the cloud. The failure of physical machines is considered by making use of the markov-chain model to do a probabilistic prediction of failure. This paper consists of two parts, one is to determine the optimum number of resource are in working condition and second is to determine effective use of these resources. We try to allocate the resources among those systems that are operational. The benefits of the system are that it allows the user to mainly predict the failure of the system and also find the optimal throughput by optimal resource allocation among those systems. We propose a general methodology and simulated the proposed methodology and illustrated with a case study.