Run-time voltage hopping for low-power real-time systems
Proceedings of the 37th Annual Design Automation Conference
Xen and the art of virtualization
SOSP '03 Proceedings of the nineteenth ACM symposium on Operating systems principles
Scheduling for reduced CPU energy
OSDI '94 Proceedings of the 1st USENIX conference on Operating Systems Design and Implementation
Queue - Virtualization
TSB: A DVS algorithm with quick response for general purpose operating systems
Journal of Systems Architecture: the EUROMICRO Journal
Proceedings of the VLDB Endowment
Communications of the ACM
Network calculus: a theory of deterministic queuing systems for the internet
Network calculus: a theory of deterministic queuing systems for the internet
Server workload analysis for power minimization using consolidation
USENIX'09 Proceedings of the 2009 conference on USENIX Annual technical conference
Accurate emulation of CPU performance
Euro-Par 2010 Proceedings of the 2010 conference on Parallel processing
Analysis of Virtualization Technologies for High Performance Computing Environments
CLOUD '11 Proceedings of the 2011 IEEE 4th International Conference on Cloud Computing
A simulator to assess energy saving strategies and policies in HPC workloads
ACM SIGOPS Operating Systems Review
Green Cloud Computing Modelling Methodology
UCC '13 Proceedings of the 2013 IEEE/ACM 6th International Conference on Utility and Cloud Computing
| Hi-index | 0.00 |
Nowadays, virtualization is present in almost all computing infrastructures. Thanks to VM migration and server consolidation, virtualization helps in reducing power consumption in distributed environments. On another side, Dynamic Voltage and Frequency Scaling (DVFS) allows servers to dynamically modify the processor frequency (according to the CPU load) in order to achieve less energy consumption. We observe that DVFS is mainly used, but still generates a waste of energy. In fact, the DVFS frequency scaling policies are based on advertised processor frequency. By default and thanks to the on demand governor, it scales up or down the processor frequency according to the current load and the different predefined threshold (up and down). However, the set of frequencies constitutes a discrete range of frequencies. In this case, the frequency required for a specific load will almost be scaled to a frequency more higher than expected, which leads to a non-efficient use of energy. In this paper, we analyze and address a way of emulating a precise CPU frequency thanks to the DVFS management in virtualized environments. We implemented and evaluated our prototype in the Xen hyper visor.