The Research On Vm Live Migration In IaaS

With the development of technology and the requirements of resources,computer architecture gradually progresses, which moves from HPC (verticalscalability), grad computing to the latest cloud computing (horizontal scalability).Cloud computing refers to both the software and hardware resources delivered asservices over the Internet from datacenters. IaaS (Infrastructure-as-a-Service) cloudwhich is from a hardware point of view allows users to boot given number ofVM(virtual machine) without maintaining how it goes. The key technology for nextgeneration of IaaS cloud services is live migration of VM, contributing to dynamicportability and mobility of VM-based services among datacenters. It allows a cleanseparation between hardware and software, and facilitates fault management,load-balancing, and low-level system maintenance.VM live migration in the cloud means migrate an entire OS while the webservice is still alive. In this case, the hardware resources are shared by both themigration process and web service process. However, when a VM is running a liveservice, it is important that its transfer operates in a manner that balances therequirements of minimizing both downtime and total migration time.In order to achieve these goals, the primary research works include the followingaspects:Analysis Clark’s live migration algorithm deeply and build a live migrationplatform based on OpenStack open source project.According queuing theory M/M/1,Model the migration bandwidth and cost function. Building an experiment to analysisthe model.Design a bandwidth provision algorithm through observing Clark’sdynamic bandwidth migration and fixed bandwidth migration, then compare newalgorithm with Clark’s algorithm through some experiments.