Research On Cloud Computing Resource Scheduling Algorithm Based On The Utilization And Load Balancing

Cloud computing enables users to access at any time via the Internet gettingvirtually unlimited computing power and a variety of information services, users ofcomputing and services readily available, easy to expand, on-demand pricing.The datacenter as a cornerstone of the implementation of cloud computing, its roleis becoming increasingly important, in the face of the situation is more complex.Increasingly large-scale, dynamic allocation and management of virtual and physicalresources are the new challenges. At present, the problems of datacenters are more andmore serious, such as the deterioration of resource utilization, the physical machine loadimbalance and high cloud datacenter power consumption problem. This paper focuseson the cloud datacenter resource scheduling, improving cloud resource utilization,making the cloud datacenters’ load balancing. This paper studies the following aspects:(1) For cloud resource scheduling, this paper analyzes the architecture of clouddatacenter, describes the main resource in cloud datacenter, gives the definition of theseresources.(2) Cloud datacenter resources for heterogeneous physical machine bring a varietyof uneven distribution of resources and resource utilization is low, we build the virtualmachine placement problem into multi-dimensional mathematical model. Based on thismathematical model, we design and implement a multi-dimensional resource virtualmachine placement algorithms based on ant colony optimization algorithm. In order toverify the performance of the algorithm, we use the open source cloud simulationplatform CloudSim implementing greedy strategy and basic ant colony algorithmplacement algorithms. The experiments show that the virtual machine placementalgorithm is verified.(3) For cloud datacenter physical machine resources load imbalance problem, wepropose a load balancing algorithm based on virtual machine migration. The loadbalancing algorithm predict multi-threshold triggering on the adjustment timing;according to the load characteristics of the virtual machine to complete the migration ofvirtual machines with the best selection and adaptation algorithm to find the virtualmachine migration target physical machine; through best load range bound noderesource utilization, so that the physical resource load balancing; according to systemload saturation, adjust the cloud datacenter physical number of machines to achieve loadbalancing and cloud datacenters energy-saving purposes.(4) Based on the cloud datacenter resource scheduling research, we analyze cloudresource scheduling simulation system’s requirements, complete the system platformmodule design and database design, according to the design of various functional modules of the system are divided into different components for development. Finally,we test cloud resource scheduling system simulation platform for verification.