Abstract
Cloud computing experiments can be conducted on simulators before they are tested on real data centre infrastructure. However, the validation of cloud system simulations is not a well-investigated topic in the literature. It can be difficult to select the suitable tools of simulation to predict infrastructure performance due to the variety of existing cloud computing simulators and their varying levels of accuracy. This paper applies an approach to evaluate any cloud simulator through a comparative validation method against micro data centre infrastructure. We have extended earlier work in evaluating cloud computing simulators by applying the method we used to evaluate CloudSim on other tools. We present the level of accuracy for the GreenCloud simulator and Mininet platforms by applying the comparative validation method and sensitivity analysis for the results on actual and simulated performance. According to our findings, GreenCloud currently does not predict the energy consumption for a micro datacenter accurately. On the other hand, Mininet shows reasonable accuracy in modelling the network performance.