The Journal of Supercomputing

, Volume 68, Issue 3, pp 1088–1112 | Cite as

Adaptive global power optimization for Web servers

  • Leonardo Piga
  • Reinaldo A. Bergamaschi
  • Mauricio Breternitz
  • Sandro Rigo


This work investigates power and performance trade-offs for Web servers on a state-of-the-art, high-density, power-efficient SeaMicro SM15k cluster by AMD. We relied on the concept of virtual power states (VPSs), a combination of CPU utilization rate to the P/C power states available in modern processors, and on our global optimization algorithm called Slack Recovery, to deploy an adaptive global power management system in a production environment. The main contributions of this paper are twofold. First, it presents the Slack Recovery algorithm deployed on a real cluster, composed of 25 SeaMicro nodes. The algorithm finds a P-state and a utilization rate for each CPU node to minimize power under a minimum performance requirement. Second, it proposes a novel mechanism to control utilization rates in each server, a key aspect on our power/performance optimization system which enables the implementation of the VPS concept in practice. Experimental results show that our Slack Recovery-based system can reduce up to 6.7 % of the power consumption when compared to policies usually deployed in SeaMicro production systems.


Power management High-density servers Web server  Power optimization Cluster 



Financial support for this study was provided by the Grant 2010/05389-5 from Sao Paulo Research Foundation (FAPESP) and AMD Research.


  1. 1.
    libpfm4 documentation. online, 2013. Accessed on 04th July 2013
  2. 2.
    Abbasi Z, Varsamopoulos G, Gupta SKS (2012) Tacoma: server and workload management in internet data centers considering cooling-computing power trade-off and energy proportionality. ACM Trans Archit Code Optim 9:2CrossRefGoogle Scholar
  3. 3.
    AMD (2012) SeaMicro SM15000 Fabric Compute Systems. Sunnyvale, CA, USAGoogle Scholar
  4. 4.
    Bergamaschi RA, Piga L, Rigo S, Azevedo R, Araujo G (2012) Data center power and performance optimization through global selection of p-states and utilization rates. Sustain Computi Inform SystGoogle Scholar
  5. 5.
    Bertini L, Leite JCB, Mossé D (2010) Power optimization for dynamic configuration in heterogeneous web server clusters. J Syst Softw 83:4CrossRefGoogle Scholar
  6. 6.
    Bianchini R, Rajamony R (2004) Power and energy management for server systems. ComputerGoogle Scholar
  7. 7.
    Brodowski D (2013) CPU frequency and voltage scaling code in the Linux(TM) kernel. Tech. rep., kernel.orgGoogle Scholar
  8. 8.
    Chase JS, Anderson DC, Thakar PN, Vahdat AM, Doyle RP (2001) Managing energy and server resources in hosting centers. In: Proceedings of the eighteenth ACM symposium on operating systems principles, SOSP ’01Google Scholar
  9. 9.
    Chen Y, Das A, Qin W, Sivasubramaniam A, Wang Q, Gautam N (2005) Managing server energy and operational costs in hosting centers. In: Proceedings of the 2005 ACM SIGMETRICS international conference on measurement and modeling of computer systems, SIGMETRICS ’05, pp 303–314Google Scholar
  10. 10.
    Cochran R, Hankendi C, Coskun A, Reda S (2011) Pack & cap: adaptive dvfs and thread packing under power caps. In: 44th annual IEEE/ACM international symposium on microarchitectureGoogle Scholar
  11. 11.
    Economou D, Rivoire S, Kozyrakis C (2006) Full-system power analysis and modeling for server environments. In: Workshop on modeling benchmarking and simulation (MOBS)Google Scholar
  12. 12.
    Elnozahy EN, Kistler M, Rajamony R (2003) Energy-efficient server clusters. In: Proceedings of the 2nd international conference on power-aware computer systems, PACS’02Google Scholar
  13. 13.
    Elnozahy M, Kistler M, Rajamony R (2003) Energy conservation policies for web servers. In: Proceedings of the 4th conference on USENIX symposium on internet technologies and systems, vol 4, USITS’03Google Scholar
  14. 14.
    Ferdman M, Adileh A, Koçberber YO, Volos S, Alisafaee M, Jevdjic D, Kaynak IC, Popescu AD, Ailamaki A, Falsafi B (2012) Clearing the clouds: a study of emerging scale-out workloads on modern hardware. In: Seventeenth international conference on architectural support for programming languages and operating systems (ASPLOS’12), pp 37–48Google Scholar
  15. 15.
    Filani D, He J, Gao S, Rajappa M, Kumar A, Shah P, Nagappan R (2008) Dynamic data center power management trends, issues, and solutions. Intel Technol JGoogle Scholar
  16. 16.
    Hackenberg D, Ilsche T, Schone R, Molka D, Schmidt M, Nagel W (2013) Power measurement techniques on standard compute nodes: a quantitative comparison. In: IEEE International symposium on performance analysis of systems and software (ISPASS), pp 194–204Google Scholar
  17. 17.
    Intel (2013) Intel 64 and IA-32 architectures software developer’s manual vol 3B. System Programming Guide, Part 2. Santa Clara, CA, USAGoogle Scholar
  18. 18.
    Isci C, Buyuktosunoglu A, Cher C, Bose P, Martonosi M (2006) An analysis of efficient multi-core global power management policies: maximizing performance for a given power budget. In: 39th annual IEEE/ACM international symposium on microarchitecture (MICRO-39 2006)Google Scholar
  19. 19.
    Kant K, Murugan M, Du DHC (2012) Enhancing data center sustainability through energy-adaptive computing. J Emerg Technol Comput Syst 8:4CrossRefGoogle Scholar
  20. 20.
    Koomey JG (2007) Estimating total power consumption by servers in the U.S. and the world. Tech. rep., Stanford UniversityGoogle Scholar
  21. 21.
    Koomey JG (2011) Growth in data center electricity use 2005 to 2010. Stanford University, Tech. repGoogle Scholar
  22. 22.
    Kusic D, Kephart JO, Hanson JE, Kandasamy N, Jiang G (2008) Power and performance management of virtualized computing environments via lookahead control. In: Proceedings of the 2008 international conference on autonomic computing, ICAC ’08Google Scholar
  23. 23.
    Leverich J, Monchiero M, Talwar V, Ranganathan P, Kozyrakis C (2009) Power management of datacenter workloads using per-core power gating. IEEE Comput Archit Lett 8(2):48–51CrossRefGoogle Scholar
  24. 24.
    Malone C, Belady C (2006) EAC & PUE: metrics to characterize IT equipment & data center energy use. In: Digital power forumGoogle Scholar
  25. 25.
    Meisner D, Sadler CM, Barroso LA, Weber W-D, Wenisch TF (2011) Power management of online data-intensive services. In: Proceedings of the 38th annual international symposium on computer architecture, ISCA ’11Google Scholar
  26. 26.
    Pallipadi V, Starikovskiy A (2006) The ondemand governor: past, present and future. In: Proceedings of Linux symposiumGoogle Scholar
  27. 27.
    Piga L, Bergamaschi R, Klein F, Azevedo R, Rigo S (2011) Empirical web server power modeling and characterization. In: IEEE international symposium on workload characterization (IISWC), 2011, p 75Google Scholar
  28. 28.
    Rajamani K, Lefurgy C (2003) On evaluating request-distribution schemes for saving energy in server clusters. In: Proceedings of the 2003 IEEE international symposium on performance analysis of systems and software, ISPASS ’03, pp 111–122Google Scholar
  29. 29.
    Rotem E, Naveh A, Rajwan D, Ananthakrishnan A, Weissmann E (2012) Power-management architecture of the intel microarchitecture code-named sandy bridge. IEEE Micro 32(2):20–27CrossRefGoogle Scholar
  30. 30.
    Schneider D (2011) Under the hood at google and facebook. online, 2011. Accessed on 20th Aug 2013
  31. 31.
    Schulz G (2009) The green and virtual data center, 1st edn. Auerbach Publications, BostonCrossRefGoogle Scholar
  32. 32.
    Shen K, Shriraman A, Dwarkadas S, Zhang X (2012) Power and energy containers for multicore servers. In: Proceedings of the 12th ACM SIGMETRICS/PERFORMANCE joint international conference on measurement and modeling of computer systems, SIGMETRICS ’12Google Scholar
  33. 33.
    Tarreau W (2013) HAProxy configuration manual version 1.5. Tech. rep., HAProxyGoogle Scholar
  34. 34.
    Vogelsang T (2010) Understanding the energy consumption of dynamic random access memories. In: Proceedings of the 2010 43rd annual IEEE/ACM international symposium on microarchitecture, MICRO ’43, pp 363–374Google Scholar
  35. 35.
    Winter JA, Albonesi DH, Shoemaker CA (2010) Scalable thread scheduling and global power management for heterogeneous many-core architectures. In: Proceedings of the 19th international conference on parallel architectures and compilation techniques, PACT ’10Google Scholar

Copyright information

© Springer Science+Business Media New York 2014

Authors and Affiliations

  • Leonardo Piga
    • 1
  • Reinaldo A. Bergamaschi
    • 1
  • Mauricio Breternitz
    • 2
  • Sandro Rigo
    • 1
  1. 1.Institute of ComputingUniversity of Campinas (UNICAMP)CampinasBrazil
  2. 2.Advanced Micro DevicesAustinUSA

Personalised recommendations