Evaluation of an RTOS on top of a hosted virtual machine system

In this paper we evaluate a virtualized RTOS by detailing its internal fine-grained overheads and latencies rather than by providing more global results from an application perspective, as it is usually the case. This approach is fundamental to analyze a mixed criticality real-time system where applications with different levels of criticality must share the same hardware with different operating systems. This evaluation allows to observe how the RTOS behaves when deployed on top of a virtual machine system and to understand what are the key features of the RTOS which impact the performance degradation.