Performance Forecast and Provisioning Optimization for IaaS Cloud
Yi Zhang, Li Dong, Shi Huiqing, Sheng Wan, Qingfeng Zhang · 2015
In this project we build up a system to improve management of IBM GCG Ring Cloud which is an internal IaaS cloud environment, i.e. Providing performance predictions and provisioning optimization for this cloud. Based on BA methods, we make this cloud environment predictable on performance of CPU, memory and network for several days in the future. We also segment these VMs depending on their performance in history and predicted days, so as to make VM resource provisioning smarter. We has finished performance prediction for all VMs and hosts, tried prototype methods for VM clustering and classification, and will continue to work on provisioning and resource optimization for clusters and data enters of this cloud.