Queueing Theory to Describe Adaptive Mathematical Models of Computational Systems with Resource Virtualization and Model Verification by Similarly Configured Virtual Server

The paper dwells upon preparing and verifying mathematical models of virtualized computational systems (CS). Today, virtualization and virtual machines are among the most efficient approaches to improving computer performance. This research is to describe, study, and verify adaptive models of virtualized CS by using servers configured similarly to such models; the researchers also seek to develop a method for making adaptive virtualization-based CS models. The research is to produce a procedure for experimental making and verification of mathematical virtualized-CS models; the procedure will create a virtual server on a host platform and monitor its performance under heavy load. Virtualization can use hardware and firmware, or software (OS level). This paper considers natural virtualization. Known models are not applicable to virtualized CS, as they cannot run comprehensive analysis to find the most efficient way of initial resource distribution and to optimize such distribution for a particular application. The mathematics this paper uses is the closed queueing networks (CQN). Experiments produced simple models for analyzing various CS structures. To be adaptive, the models use triggers that track and adjust the processing channel power in individual queueing systems (QS) to a particular application. The conclusions present the findings of this research. Experiments prove the obtained results reliable and usable as a flexible tool for studying the virtualization properties when structuring a CS. This knowledge could be of use for businesses interested in optimizing the server configuration for their IT infrastructure, e.g. for testing, remote offices, etc., since renting a virtual server is cheaper than buying one.