Design And Implementation Of Iaas Platform Resource Management Scheme

As a new distributed application system, cloud computing can integrate the resources of the single data center or multiple data center, and provide users with infrastructure (Infrastructure as a Service, IaaS), platform (Platform as a Service, PaaS) or software (Software as a Service, SaaS) via Internet. This article will focus on the design and implementation of IaaS platform resource management solution.The application of IaaS could provide flexibility IT infrastructure for the enterprise with high quality performance and scalability. However, it also brings some prominent issues:first, the introduction of virtualization technology has brought more complicated infrastructure management, including the integration of virtual and physical resources and the unified management of heterogeneous virtual resources; second, it is necessary to optimize the scheduling strategy to achieve the reasonable scheduling and load balancing of the cloud resources.In order to solve the two issues stated, this article anylyzes and discusses the basic principles and architecture of cloud computing and virtrualization, the general architecture model of IaaS platform, and the private cloud solution--ISF of Platform Computing Company. Based on ISF, the article has designed and implemented a IaaS platform resource management solution. This design consists of two layers: resources abstraction layer which solves the integration of physical and virtual resources, and the resources scheduling layer which provides the scheduling strategy.Resource abstraction layer has provided a simple interface for IaaS platform to manage the heterogeneity of IaaS platform management infrastructure. Under the management of the resource abstraction layer, heterogeneous virtual resources and physical resources could be scheduled and allocated by the consistent management interface. On the basis of resource abstraction layer, the resource scheduling layer provides a variety of resource scheduling strategies which are suitable to the heterogeneous resources, and optimize resource utilization, balance the loading and with high availability. Finally, the solution and scheduling strategy are tested by a set of test cases, and the method proposed in this article is proved to solve the issues of resources abstraction and resource utilization well.