Research And Implementation Of Private Cloud Platform For Virtual Resource Self-adaptive

With the rapid development of information technology such as computers andInternet, informatization has become an effective means to improve core competence.At present, the majority of enterprises have completed the construction ofinformatization and adopted an information service model with a data center businessoperation platform. Due to the purpose of ensure information services run smoothly,these enterprises often plan data center capacity in accordance with the peak valuewhen built data centers,, infrastructure resources are not fully utilized. As the thirdwave of IT Cloud computing has walked into our lives, it brings a fundamental changein human life, production methods and business models, and then it becomes the focusof attention of the whole society. Cloud computing can not only improve serverutilization, offer high-performance computing and mass data storage, but also integrateand use the existing IT infrastructure, which reduces the capital investment ofenterprises in the construction and maintenance of the data center, so it becomes aneffective method to solve a series of problems encountered in the construction ofinformatization.This paper ensured the needs of a private cloud platform and divided it intoinfrastructure layer, cloud middleware layer and cloud portal layer functionally on thebasis of the analysis of the problems of the research status of private cloud platformand the existing problems of data center This paper designed the open sourceconstruction scheme and construction architecture of a private cloud platform. On theinfrastructure layer, this paper completed the design and realization of a private cloudplatform on the basis of KVM and Openstack.On the cloud middleware layer, thispaper designed and implemented resource monitoring, resource prediction, virtualresources self-adaption, user management and cloud storage to meet the applicationrequirements of enterprises. On the cloud portal layer, this paper completed a customdevelopment in OpenStack Horizon to offer a visual User interface for adoption ofcloud intermediate layer functions The self-adaptive virtual resources is the corefunction of the private cloud platform which is designed into three processing subprogram: The visual machine is used to adjust the multi-granularity of virtualmachines according to the load state of the virtual machine and its cluster, andimproving the resource utilization rate while ensuring business continuity by means ofdynamic configuration of virtual machine resources, virtual machine live migrationand the business load regulation; The server load optimization is used to divide serverinto a different state, and optimizing the sever when its load is too large or too small，improving server performance; The cloud platform load optimization is used to adjustall the server load in as a whole to achieve the best as far as possible and improveserver resource utilization.This paper studied and realized a private cloud platform to offer self-adaptivevirtual resources and cloud storage services, and also can fulfil the demands in datacenter construction of current enterprises and institutions. The application,implementation and test of the cloud platform indicating that the private cloud platformsolution is feasible.