CRAM: Efficient Dynamic Resource Scheduling in Cloud Computing

Abstract: -Cloud computing has emerged as the default paradigm for a variety of fields especially considering the resources and infrastructures consumption in case of distributed access. The solution has however placed a lot of emphasis of a cloud server with variety of demands of which quality of service reminds a paramount strategy. There are a lot of strategies is in place for this quality cost which is regulated by service level agreement [SLA].An SLA is an agreement between client and server which when violated will impose penalties for the infringement or violation performance evaluation place a key role allowing system managers to evaluate the effects of different resource management strategies on the data center functioning and to predict the corresponding cost/benefits. In order to deal with very large systems composed of hundreds or thousands of resources. The system should allow to easily implementing different strategies and should have policies to represent different working conditions. Keeping this is mind a rewards scheme called as SRNS-Stochastic Reward Nets is utilized which is dynamic in nature to the status of the requests made and job allocated. The proposed model is scalable enough to represent system composed of thousands of resources and it makes possible to represent both physical and virtual resources exploiting cloud specific concepts such as the infrastructure elasticity. The resources offered by other public cloud system through a sharing and paying model is also controlled in term of performance metrics like utilization, responsiveness and load burst. This model is analytical in nature unlike the existing approaches which are assumptive and simulation based in nature