Correlationship-oriented Resource Scheduling In Cloud Computing

In the age of the Internet, cloud computing is a way to turn a large amount of computing resources as needed which enables users to run a large number of emerging model of computing tasks. Scheduler of the cloud computing platform which is the core of integration modules plays an important role in optimizing resources, improving task performance and reducing risk platform. As the task which is worked on the cloud computing platform becomes more and more complex, more and more diverse, the role of the scheduler becomes more and more important in cloud computing platform. As computing tasks Load increases, cloud computing platform for virtual resource management becomes a new challenge for cloud computing platform. The present cloud computing platform for correlationship tasks scheduling there are low rates of resource utilisation, inequitable distribution led to uneven loading, thereby affecting the performance of cloud computing platforms.In order to Improve resource utilization, resource allocation fairness cloud computing platforms, as well as load balancing problems, we presents correlationship-baseddomainant-fairness algorithm(CBDRF). Major research efforts in the following areas.(1) In the research of management for cloud computing resources, combined with the structure of the cloud computing platform, focuses on the characteristics of resource allocation, scheduling model platform, and the platform scheduler.(2) To solve the problem of the correlationship tasks in the allocation of resources for cloud computing platform of low resource utilization and insufficient equity. We use graph theory model combined algorithm for min-max principle and the DRF, to model the resource scheduling problems with cloud computing platforms. Based on this model we design and implement CBDRF scheduling algorithm which based on DRF algorithm and graph algorithms which is able to schedule the correlationship tasks with balance of fairness and efficiency, and we effectively improve the utilization of resources.(3) Based on the realization of CBDRF algorithm and combining cloud computing schedule research present situation and needs. we complete implementation of scheduler and tests the algorithm on the Openstack platform.