Uncertainty analysis of virtualized systems with live migration
J. Zhang, Jinke Zheng, H. Okamura, Tadashi Dohi · IET conference proceedings. · 2023
Benefiting from the advantages of virtualization technology, virtualized systems have been widely studied and adopted in enterprise systems, aimed at reducing resource usage and enhancing system reliability. An important feature of virtualized systems is the virtual machine (VM) live migration, which brings convenience to transfer running VMs from one physical server to another without service interruption. To achieve a high available/reliable virtualized system, it is crucial to evaluate its system availability/reliability. The model-based approach is commonly used and proved effective. However, it is well-known that the model parameters typically involve estimation error due to the limitation of data samples or expert's knowledge, so-called uncertainty. In this paper, we analyze the uncertainty propagation in the availability model of a virtualized system with live migration through a momentbased approximation method. A numerical experiment is conducted to quantify the impact of the parameter uncertainty on the system availability.