Towards Fair And Efficient Virtual Machine Scheduling Mechanism In Cloud Computing

Through the usage of virtualization technology to reorganize the underlying physical resources and virtualizing them, cloud computing is forming a huge pool of resources, computing center will provide these resources in a flexible and dynamic type of service to the user, which enable the users to obtain the required computing services(computing, network,storage and security, etc) in an extremely vulnerable way. Resource scheduling module lies in the core position in cloud computing controlling, and is the key component to large-scale applications and cloud computing to improve system performance.At present, the independency of different nodes,utilization of single node or cluster load are the key points of the study focused, ignoring the relationship between the virtual machines. From the task request’s point of view,the scheduler reduces the possibility of collaborative demand that cooperate between virtual machines, efficiently, thus affecting these applications that need to be calculated in collaboration with the quality of the service. From the cluster perspective, service provider expects fair allocation of these resources to enhance the system’s load- balancing capability and reduce system jitter.To address these defects above, a virtual machine shcheduling strategy-Balancing Efficiency and Fairness in Virtual System with Network Aware Allocation, namely BEFNAA, has been proposed in this paper. In addition, the scheduler in Collaborative Computing Platform has been designed and realized. A two-step scheduling mechanism has been designed in BEF-NAA, which generates the allocation sets that meet the fariness characteristics based on cluster resources consideration, firstly. Next in importance, communication requirements between nodes are taken into account to merge. BEF-NAA considers both the cluster and the tasks, aming to achieve the process of allocation and merging with a more reasonable configuration, thus achieving a integrated optimization on two aspects.Testing results show that BEF-NAA addresses the communication delay between nodes, reducing the transfer time when deploying the computing framework and fetching the image copys on the premise of ensuring the cluster loads. BEF-NAA can provide a more feasible strategy support for cloud computing and new optimizing ideas for vm-based scheduling.