Resource Scheduling System Based On Hybrid Cloud

With the rapid development of Internet technology,cloud computing technology has been widely recognized and used by more and more industries.At present,the cloud computing platform is mainly divided into public cloud and private cloud.Public cloud is mainly face to the general people that pay attention to lower cost and more convenient services,but it cannot provide a good guarantee for the security of application and data.Private cloud mainly satisfies interior requirement of enterprise and university,and it focus on data security and quality assurance.Due to the limitation of cost and scale,it would reduce the quality of service in private when the number of applications surge in a particular time.Hybrid cloud has both merits of data security and low cost,so it has been seen as mainstream model of future cloud computing system.In this paper,we use OpenStack to build a private cloud platform in the university,and we also implement a hybrid cloud resource scheduling system based on the existing public clouds and the other two private cloud in different universities.In hybrid cloud system,resources can be scheduled more reasonable according to different resource request by users,and we design a series of scheduling policies including price,performance,security,availability,balance,etc.This system can not only be used for interior resource scheduling to reduce average delay of data retrieving in a private cloud but also used for resource scheduling across private cloud and public cloud.When resources in a private cloud are insufficient,it would automatic extend request to public cloud,and when high security applications need to be deployed,it would automatic schedule those to the private cloud platform.Based on the real private and public cloud platform,we deploy many practical applications in the hybrid cloud system.It finally comes to the conclusion that it can reach more than 90% accuracy of resource scheduling across different cloud,and achieve 20% lower average delay in a private cloud inside.