Advertisement

Analyzing the power consumption behavior of a large scale data center

  • Kashif Nizam KhanEmail author
  • Sanja Scepanovic
  • Tapio Niemi
  • Jukka K. Nurminen
  • Sebastian Von Alfthan
  • Olli-Pekka Lehto
Special Issue Paper
  • 128 Downloads

Abstract

The aim of this paper is to illustrate the use of application and system level logs to better understand scientific data center behavior and energy-spending. Analyzing a data center log of 900 nodes (Sandy Bridge and Haswell), we study node power consumption and describe approaches to estimate and forecast it. Our results include methods to cluster nodes based on different vmstat and RAPL measurements as well as Gaussian and GAM models for estimating the plug power consumption. We also analyze failed jobs and find that non-successfully terminated jobs consume around 40% of computing time. While the actual numbers are likely to vary in different data centers at different times, the purpose of the paper is to share ideas of what can be found by statistical and machine learning analysis of large amount of log data.

Keywords

RAPL Energy modeling Energy efficiency Data center log analysis 

Notes

Acknowledgements

Author Kashif Nizam Khan would like to thank Nokia Foundation for a grant which helped to carry out this work.

References

  1. 1.
    Taito supercluster. https://research.csc.fi/csc-s-servers/taito. Accessed 17 Marc 2017
  2. 2.
    Bircher WL, John LK (2012) Complete system power estimation using processor performance events. IEEE Trans Comput 61(4):563–577.  https://doi.org/10.1109/TC.2011.47 MathSciNetCrossRefzbMATHGoogle Scholar
  3. 3.
    Borghesi A, Bartolini A, Lombardi M, Milano M, Benini L (2016) Predictive modeling for job power consumption in HPC systems. Springer, Cham, pp 181–199Google Scholar
  4. 4.
    Breiman L (2001) Random forests. Mach Learn 45(1):5–32CrossRefzbMATHGoogle Scholar
  5. 5.
    Dayarathna M, Wen Y, Fan R (2016) Data center energy consumption modeling: a survey. IEEE Commun Surv Tutor 18(1):732–794CrossRefGoogle Scholar
  6. 6.
    Economou D, Rivoire S, Kozyrakis C, Ranganathan P (2006) Full-system power analysis and modeling for server environments. In: International symposium on computer architecture-IEEEGoogle Scholar
  7. 7.
    Ge R, Feng X, Song S, Chang HC, Li D, Cameron KW (2010) Powerpack: energy profiling and analysis of high-performance systems and applications. IEEE Trans Parallel Distrib Syst 21(5):658–671CrossRefGoogle Scholar
  8. 8.
    Hackenberg D, Schöne R, Ilsche T, Molka D, Schuchart J, Geyer R (2015) An energy efficiency feature survey of the Intel Haswell processor. In: 2015 IEEE international parallel and distributed processing symposium workshop, pp. 896–904.  https://doi.org/10.1109/IPDPSW.2015.70
  9. 9.
    Hirki M, Ou Z, Khan KN, Nurminen JK, Niemi T (2016) Empirical study of the power consumption of the x86-64 instruction decoder. In: USENIX workshop on cool topics on sustainable data centers (CoolDC 16). USENIX Association, Santa Clara, CAGoogle Scholar
  10. 10.
    Intel: Intel 64 and IA-32 Architectures Software Developer’s Manual Volume 3 (3A, 3B & 3C): System Programming Guide (2014)Google Scholar
  11. 11.
    Khan KN, Ou Z, Hirki M, Nurminen JK, Niemi T (2016) How much power does your server consume? Estimating wall socket power using RAPL measurements. Comput Sci Res Dev 31(4):207–214CrossRefGoogle Scholar
  12. 12.
    Kohonen T (1998) The self-organizing map. Neurocomputing 21(1):1–6MathSciNetCrossRefzbMATHGoogle Scholar
  13. 13.
    Möbius C, Dargie W, Schill A (2014) Power consumption estimation models for processors, virtual machines, and servers. IEEE Trans Parallel Distrib Syst 25(6):1600–1614CrossRefGoogle Scholar
  14. 14.
    Molka D, Hackenberg D, Schöne R, Müller MS (2010) Characterizing the energy consumption of data transfers and arithmetic operations on x86-64 processors. In: International conference on green computing, pp 123–133Google Scholar
  15. 15.
    Podzimek A, Bulej L, Chen LY, Binder W, Tuma P (2015) Analyzing the impact of cpu pinning and partial cpu loads on performance and energy efficiency. In: 2015 15th IEEE/ACM international symposium on cluster, cloud and grid computing, pp 1–10.  https://doi.org/10.1109/CCGrid.2015.164
  16. 16.
    Shehabi A, Smith S, Horner N, Azevedo I, Brown R, Koomey J, Masanet E, Sartor D, Herrlin M, Lintner W (2016) United states data center energy usage report. Lawrence Berkeley National Laboratory, Berkeley, California. LBNL-1005775, p 4Google Scholar
  17. 17.
    Zhai Y, Zhang X, Eranian S, Tang L, Mars J (2014) HaPPy: hyperthread-aware power profiling dynamically. In: 2014 USENIX annual technical conference (USENIX ATC 14), pp 211–217. USENIX Association, Philadelphia, PAGoogle Scholar

Copyright information

© Springer-Verlag GmbH Germany, part of Springer Nature 2018

Authors and Affiliations

  1. 1.Helsinki Institute of PhysicsHelsinkiFinland
  2. 2.Aalto UniversityEspooFinland
  3. 3.CSC - IT Center for ScienceEspooFinland

Personalised recommendations