Optimizing I/O Intensive Domain Handling in Xen Hypervisor for Consolidated Server Environments

Conference paper
First Online:
Part of the Lecture Notes in Computer Science book series (LNCS, volume 9663)

Abstract

Consolidation of servers through virtualization, facilitated by the use of hypervisors, allows multiple servers to share a single hardware platform. Xen is a widely preferred hypervisor, mainly, due to its dual virtualization modes, virtual machine migration support and scalability. This paper involves an analysis of the virtual CPU (vCPU) scheduling algorithms in Xen, on the basis of their performance while handling compute intensive or I/O intensive domains in virtualized server environments. Based on this knowledge, the selection of CPU scheduler in a hypervisor can be aligned with the requirements of the hosted applications. We introduce a new credit-based vCPU scheduling strategy, which allows the vCPUs of I/O intensive domains to supersede other vCPUs, in order to favor the reduction of I/O bound domain response times and the subsequent bottleneck in the CPU run queue. The results indicate substantial improvement of I/O handling and fair resource allocation between the host and guest domains.

Keywords

Xen hypervisor Server consolidation Virtual machine monitor (VMM) CPU scheduling 
This is a preview of subscription content, log in to check access.

Notes

Acknowledgement

The authors would like to thank the anonymous reviewers for their comments and suggestions. This work is partly supported by NSERC Grant CRDPJ 445731-12.

References

  1. 1.
    Natural Resources Defense Council (NRDC). Data Center Energy Assessment Report (2014). https://www.nrdc.org/energy/files/data-center-efficiency-assessment-IP.pdf
  2. 2.
    Hussain, T., Habib, S.: A redesign methodology for storage management virtualization. In: IFIP/IEEE International Symposium on Integrated Network Management (IM), pp. 676–679 (2013)Google Scholar
  3. 3.
    Barham, P., Dragovic, B., Fraser, K., et al.: Xen and the art of virtualization. In: 19th ACM Symposium on Operating Systems Principles, pp. 164–177 (2003)Google Scholar
  4. 4.
    Duda, K.J., Cheriton, D.R.: Borrowed-virtual-time (BVT) scheduling: supporting latency-sensitive threads in a general-purpose scheduler. In: 17th ACM SIGOPS Symposium on Operating Systems Principles, pp. 261–276 (1999)Google Scholar
  5. 5.
    Leslie, I.M., McAuley, D., Black, R., Roscoe, T., Barham, P., Evers, D., Fairbairns, R., Hyden, E.: The design and implementation of an operating system to support distributed multimedia applications. IEEE J. Sel. Areas Commun. 14(7), 1280–1297 (1996)CrossRefGoogle Scholar
  6. 6.
    Jansen, P.G., Mullender, S.J., Havinga, P.J., Scholten, H.: Lightweight EDF scheduling with deadline inheritance. University of Twente (2003). http://doc.utwente.nl/41399/
  7. 7.
    VanderLeest, S.H.: ARINC 653 hypervisor. In: 29th IEEE/AIAA Digital Avionics Systems Conference (DASC), pp. 5.E.2-1–5.E.2-20 (2010)Google Scholar
  8. 8.
  9. 9.
  10. 10.
    IOzone Filesystem Benchmark. http://www.iozone.org/
  11. 11.
  12. 12.
  13. 13.
  14. 14.
    Uddin, M., Rahman, A.: Energy efficiency and low carbon enabler green IT framework for data centers considering green metrics. Renew. Sustain. Energy Rev. 16(6), 4078–4094 (2012)CrossRefGoogle Scholar
  15. 15.
    Cheng, K., Bai, Y., Wang, R., Ma, Y.: Optimizing Soft real-time scheduling performance for virtual machines with SRT-Xen. In: 15th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing (CCGrid), pp. 169–178 (2015)Google Scholar
  16. 16.
    Qu, H., Liu, X., Xu, H.: A workload-aware resources scheduling method for virtual machine. Int. J. Grid Distrib. Comput. 8(1), 247–258 (2015)CrossRefGoogle Scholar

Copyright information

© Springer International Publishing Switzerland 2016

Authors and Affiliations

  1. 1.School of Electrical Engineering and Computer ScienceUniversity of OttawaOttawaCanada

Personalised recommendations