Evaluation of the overheads and latencies of a virtualized RTOS

The virtualization technique has become a popular trend in the domain of real-time embedded systems. For example, in the automotive industry, practitioners are currently considering the idea of using such technique to run simultaneously the AUTOSAR real-time operating system (RTOS) for real-time programming, and the Linux-GENIVI operating system to support in-vehicle infotainment applications on the same Electronic Control Unit. However, running a real-time operating system inside a virtual machine instead of a bare metal hardware may have a significant cost in terms of latencies and overheads. This raises the question of the order-of-magnitude of a virtualized RTOS's overheads and latencies in comparison to a single RTOS running on a real machine. In this paper, an evaluation of a virtualized RTOS's latencies and overheads is proposed in order to observe how virtualization impacts these timing characteristics. Our experiments demonstrate that the overheads and latencies of a virtualized RTOS compete with those obtained in the same RTOS running on bare-metal hardware.

[1]  Thomas Nolte,et al.  Evaluation of Real-Time Performance in Virtualized Environment , 2011 .

[2]  Theodore P. Baker,et al.  A Comparison of Global and Partitioned EDF Schedulability Tests for Multiprocessors TR-051101 , 2005 .

[3]  Jun Zhang,et al.  Performance Tuning Towards a KVM-Based Low Latency Virtualization System , 2010, 2010 2nd International Conference on Information Engineering and Computer Science.

[4]  Fabrice Bellard,et al.  QEMU, a Fast and Portable Dynamic Translator , 2005, USENIX ATC, FREENIX Track.

[5]  Björn B. Brandenburg,et al.  LITMUS RT : A Status Report ∗ , 2007 .

[6]  Otto-Hahn-Ring Towards Linux as a Real-Time Hypervisor , 2009 .

[7]  A. Kivity,et al.  kvm : the Linux Virtual Machine Monitor , 2007 .

[8]  James H. Anderson,et al.  Scheduling and locking in multiprocessor real-time operating systems , 2011 .

[9]  James E. Smith,et al.  Virtual machines - versatile platforms for systems and processes , 2005 .

[10]  Kai Chen,et al.  Performance analysis towards a KVM-Based embedded real-time virtualization architecture , 2010, 5th International Conference on Computer Sciences and Convergence Information Technology.

[11]  Insup Lee,et al.  Realizing Compositional Scheduling through Virtualization , 2012, 2012 IEEE 18th Real Time and Embedded Technology and Applications Symposium.

[12]  Tao Ding,et al.  Scheduling Policy Optimization in Kernel-Based Virtual Machine , 2010, 2010 International Conference on Computational Intelligence and Software Engineering.