Research And Implementation Of Severaltechnologies For Virtual Machine Deployment

The emergence and developmentof cloud computinghas turned the traditional computation unit consists of physical machines into virtual machines.Applications used to run on physical machines now run on virtual machines. Virtual machine is hosted on a physical machine. Virtual machine deployment has an important impact on resource utilization, virtual machines performance and the load balance of the entire cloud environment, so virtual machine deployment is a key issue within cloud computing.How to study and solve the virtual machine deployment is an import ant content in current engineering practice.Mainly work of this thiesis is as follows:First,Put forward three methods to create virtual machine images. The first methodis to createvirtual machine images with virtualization software;The second method is to recompile operating system source code using the LFS technology; The third method is to use the kernel cutting technology to cut the kernel functions whichis not necessary for users.Though these methods only required software is compiled and unnecessary software is not compiled,so we can create virtual machine images with small size,high performance,occupy less resource, as well as meet the demand of users.Second,Design and implement an imageservice model.Image service model is a cluster made up of a management node and several image storage nodes.A larger virtual machine image is divided into several smaller image data blocksand These image data blocks are stored in the ImageServer, and each image data block is stored redundancy, thereby the host machine can download image data blocks from different image storage server parallelly and then merged these image data blocks into a complete image. This method can effectively reduce the image transmission time and as well avoid performance bottlenecks of one single image storage server.Third,put forward a virtual machine deployment algorithm based on weight,which can balance the three target of virtual machine deployment in cloud environment: load balancing, virtual machine pe rformance, resource utilization.