Труды Института системного программирования РАН (Oct 2018)
Software performance estimation in a virtualized environment based on atomic tests
Abstract
Virtual Machine (VM) environments are experiencing a resurgence of interest for diverse uses including server consolidation and shared hosting. An application performance in a virtual machine environment can differ markedly from its performance in a non-virtualized environment because of interactions with the underlying virtual machine monitor and other virtual machines. In this paper, I describe a general approach for estimating the resource requirements of applications when they are transferred to a virtual environment. The core principle of this approach is splitting complex workload into a combination of smaller tasks and replacing these tasks with synthetic atomic tests. Performance evaluation of atomic tests on native hardware and in the virtual machine allows us to define virtualization overhead for the given platform.