Research On Resource Assessment Model In Cloud Computing

Cloud computing has become the research focus in industrial and academic area, and one of the most promising models of business computing. Cloud computing, which relies on the mature virtualization technology, and developed on the basis of distributed computing and grid computing, is aiming to providing reliable, personalized, and Qos(Quality of service) guaranteed service. Cloud computing allows users to get services just through Web browser, by simply paying the fee correspondingly. In the cloud computing environment, computing resources can be allocated and released dynamically, while users do not need to purchase the expensive servers and storage devices. All resources available to users are in the form of on-demand services, such as Infrastructure as a Service (IaaS), Platform as a service(PaaS), Software as a service(SaaS), and so on.Task scheduling has always been one of the key problems to be solved in distributed computing system, while there are some new features in service scheduling and resource allocation in Cloud Computing. The scheduling objectives are Qos-oriented, which do not only focus on the efficiency of task complement, but also the payment of resource and the system usage.To settle these problems, firstly the existing scheduling mechanisms and task scheduling algorithms are studied in this paper, and then a scheduling mechanism with resource assessment is designed, which schedules tasks according to the assessment result. This model, which uses loosely coupled distributed negotiation mechanism, has good flexibility and scalability.This model uses virtual resources as the assessment objects, including computing resources, storage resources and communications resources. Traditional best-effort based task scheduling algorithms often take completion time as the scheduling object, while in cloud computing, service quality is taken into account as the object, including the service capabilities, service costs and service reliability. What's more, to improve the system load balancing, the resource integrated load is considered. The resource assessment algorithm adopts the concept of approaching the ideal solution, which judges the evaluation value of candidates by the similarity to the ideal service.