Research On The Optimized Scheduling Of Energy-aware Vtrtual Machine Network In Heterogeneous Cloud Environment

Most data centers under cloud computing technical environment are gradually transferring into a cloud service mode by using virtualization technology,so the virtual machine network running under the heterogeneous cloud environment is becoming more complicated.To satisfy the increasing requirements of users for cloud computing requirement,the quantity and scale of the cloud data center as a key carrier are becoming larger as each day goes by.The cloud data center operators are facing severe problems such as high energy consumption and high cost.Differences in mapping from VM to physical servers will lead to different energy consumption at the data center under a heterogeneous cloud environment,so optimized scheduling of VM network with energy consumption awareness becomes the key problem in resource management of cloud data center.The cloud data center administrators are confronted with the problem as to measure energy consumption at the whole data center,perceive energy consumption at the cloud data center and efficiently manage energy consumption at the heterogeneous cloud data center,to satisfy placement and operation of the VM network at the data center with minimal energy consumption,and to adjust operation of the whole data center in time according to the dynamic change of VMs during operation,with the aim of reducing energy consumption of the data center and service cost.Therefore,this paper studies optimized scheduling of the VM network of energy consumption awareness under the heterogeneous cloud environment.First of all,for the energy consumption awareness of the heterogeneous cloud data center,this paper designs a measurement method for energy consumption under the heterogeneous cloud environment;establishes an energy consumption model of heterogeneous cloud data center,including VM energy consumption model,physical machine energy consumption model and network equipment energy consumption model;measures and predicts infrastructure energy consumption under an heterogeneous cloud environment to support optimized placement and scheduling of the VM network with energy consumption awareness.Furthermore,for optimized placement of VM network with energy consumption awareness,this paper establishes the problem model for optimized placement of VM network;solves the VM network placement problem by using the immune genetic algorithm and improved algorithm based on minimal cut and best adaptation;compares frequent placement algorithms at existing data center and some heuristic algorithms;and analyzes solution quality of the immune genetic algorithm and improved algorithm based on minimal cut and best adaptation in VM network placement problem.Additionally,for dynamic migration problem of VM network with energy consumption awareness,this paper counts history access habits of users,analyzes user's behavior characteristics,sets up a user's behavior characteristic model,establishes characteristic model via the VM load prediction,puts forward problem description,problem model and algorithm step of dynamic migration algorithms of VM network based on user's behavior analysis,and analyzes it via comparative test.Finally,based on theoretical research on optimized placement and dynamic migration of the VM network,this paper designs and develops the resource management platform of cloud data center with energy consumption awareness,designs the architecture of the resource management platform of a cloud data center with energy consumption awareness,function modules,database and key algorithms,and implements two kernel modules of the platform.