ABSTRACT
Current capacity planning practices based on heavy over-provisioning of power infrastructure hurt (i) the operational costs of data centers as well as (ii) the computational work they can support. We explore a combination of statistical multiplexing techniques to improve the utilization of the power hierarchy within a data center. At the highest level of the power hierarchy, we employ controlled underprovisioning and over-booking of power needs of hosted workloads. At the lower levels, we introduce the novel notion of soft fuses to flexibly distribute provisioned power among hosted workloads based on their needs. Our techniques are built upon a measurement-driven profiling and prediction framework to characterize key statistical properties of the power needs of hosted workloads and their aggregates. We characterize the gains in terms of the amount of computational work (CPU cycles) per provisioned unit of power Computation per Provisioned Watt (CPW). Our technique is able to double the CPWoffered by a Power Distribution Unit (PDU) running the e-commerce benchmark TPC-W compared to conventional provisioning practices. Over-booking the PDU by 10% based on tails of power profiles yields a further improvement of 20%. Reactive techniques implemented on our Xen VMM-based servers dynamically modulate CPU DVFS states to ensure power draw below the limits imposed by soft fuses. Finally, information captured in our profiles also provide ways of controlling application performance degradation despite overbooking. The 95th percentile of TPC-W session response time only grew from 1.59 sec to 1.78 sec--a degradation of 12%.
- Murali Annavaram, Ed Grochowski, and John Shen. Mitigating Amdahl's Law through EPI Throttling. In Proceedings of the International Symposium on Computer Architecture (ISCA), 2005. Google ScholarDigital Library
- Paul Barham, Boris Dragovic, Keir Fraser, Steven Hand, Tim Harris, Alex Ho, Rolf Neugebauer, Ian Pratt, and Andrew Warfield. Xen and the art of virtualization. In SOSP '03: Proceedings of the nineteenth ACM symposium on Operating systems principles, New York, NY, USA, 2003. ACM. Google ScholarDigital Library
- R. Boorstyn, A. Burchard, J. Liebeherr, and C. Oottamakorn. Statistical Service Assurances for Traffic Scheduling Algorithms. In IEEE Journal on Selected Areas in Communications, 18:12, pages 2651--2664, December 2000. Google ScholarDigital Library
- Shiva Chaitanya, Bhuvan Urgaonkar, and Anand Sivasubramaniam. QDSL: QoS-aware Systems with Differential Service Levels. In Proceedings of the ACM Conference on Measurement and Modeling of Computer Systems (SIGMETRICS), Annapolis, MD, June 2008. Google ScholarDigital Library
- J. Choi, S. Govindan, B. Urgaonkar, and A. Sivasubramaniam. Profiling, Prediction, and Capping of Power Consumption in Consolidated Environments. In IEEE International Symposium on Modeling, Analysis, and Simulation of Computer and Telecommunication Systems, 2008.Google ScholarCross Ref
- C. Clark, K. Fraser, Steven Hand, J. Hansen, E. Jul, C. Limpach, I. Pratt, and A. Warfield. Live Migration of Virtual Machines. In Proceedings of the 2nd Symposium on Networked Systems Design and Implementation (NSDI'05), May 2005. Google ScholarDigital Library
- D. Clark. Power Hungry Computers Put Data Centers in Bind. The Wall Street Journal (Online), November 2005. http://hightech.lbl.gov/DCTraining/docs/wsjon-data-ctr-power.pdf.Google Scholar
- Commercial Circuit Breakers. http://circuit--breakers.carlingtech.com/all circuits.asp.Google Scholar
- Dell Power Calculator, May 2008. http://www.dell.com/content/topics/topic.aspx/global/products/pedge/topics/en/configcalculator?c=us\&cs=555\&l=en\&s=biz.Google Scholar
- Dell SC1425, Dec 2005. http://www.dell.com/downloads/global/products/pedge/en/sc1425/specs.pdf.Google Scholar
- X. Fan, W.-D. Weber, and L. A. Barroso. Power Provisioning for a Warehouse-Sized Computer. In Proceedings of the Thirty Fourth Annual International Symposium on Computer Architecture, 2007. Google ScholarDigital Library
- W. Felter, K. Rajamani, C. Rusu, and T. Keller. A Performance-Conserving Approach for Reducing Peak Power Consumption in Server Systems. In 19th Annual International conference on Supercomputing (ICS), June 2005. Google ScholarDigital Library
- Gartner 2007. http://www.globalactionplan.org.uk/upload/resource/Full-report.pdf.Google Scholar
- S. Heins. Divorcing Electricity Sales from Profits Creates Win-Win for Utilities and Customers, Energy Pulse, Sept. 2006. http://www.energypulse.net/centers/article/article display.cfm?a id=1342.Google Scholar
- J. Hellerstein, F. Zhang, and P. Shahabuddin. A Statistical Approach to Predictive Detection. Computer Networks, January 2000. Google ScholarDigital Library
- HP Power Manager. http://h18013.www1.hp.com/products/servers/management/ilo/power-regulator.html.Google Scholar
- IBM Energy Manager. http://www-03.ibm.com/press/us/en/pressrelease/22551.wss.Google Scholar
- IBM Workload Estimator. http://www-304.ibm.com/systems/support/tools/estimator/index.html.Google Scholar
- R. Iyer, V. Tewari, and K. Kant. Overload Control Mechanisms for Web Servers. In Workshop on Performance and QoS of Next Generation Networks, 2000.Google Scholar
- Marianne Lavelle. Conservation Can Mean Profits for Utilities: States are changing the rules of the game so that it pays power companies not to expand, USNews and World Report, May 2008. http://www.usnews.com/.Google Scholar
- C. Lefurgy, X. Wang, and M. Ware. Server-Level Power Control. In ICAC '07: Proceedings of the Fourth International Conference on Autonomic Computing, Washington, DC, USA, 2007. IEEE Computer Society. Google ScholarDigital Library
- R. Nathuji and K. Schwan. Virtualpower: Coordinated power management in virtualized enterprise systems. In 21st ACM Symposium on Operating Systems Principles (SOSP'07), 2007. Google ScholarDigital Library
- New York Times Article. Hiding in plain sight, Google Seek More Power, June 2006. http://www.nytimes.com/2006/06/14/technology/14search.html?pagewanted=2.Google Scholar
- K. Park and W. Willinger. Self-Similar Network Traffic and Performance Evaluation. Wiley-Interscience, John Wiley and Sons, Inc., 2000. Google ScholarDigital Library
- R. Raghavendra, P. Ranganathan, V. Talwar, Z. Wang, and X. Zhu. No Power Struggles: Coordinated multi-level power management for the data center. In 13th International Conference on Architectural Support for Programming Languages and Operating Systems, March 2008. Google ScholarDigital Library
- L. Ramos and R. Bianchini. C-Oracle: Predictive thermal management for data centers. In Proceedings of the Fourteenth International Symposium on High-Performance Computer Architecture (HPCA'08), February 2008.Google ScholarCross Ref
- P. Ranganathan, P. Leech, D. Irwin, and Jeff Chase. Ensemble-level Power Management for Dense Blade Servers. In Proceedings of the International Symposium on Computer Architecture (ISCA), June 2006. Google ScholarDigital Library
- Raritan Inc. 20Amp PDU Model, May 2008. http://www.raritan.com/products/power-management/Dominion-PX/DPCR20-20/.Google Scholar
- Neil Rasmussen. Electrical Efficiency Modeling for Data Centers. In APC White Paper #113, 2007.Google Scholar
- B. C. Smith, J. F. Leimkuhler, and R. M. Darrow. Yield Management at American Airlines. In Interfaces, 22:1, pages 8--31, Jan 1992.Google Scholar
- W. Smith. TPC-W: Benchmarking An Ecommerce Solution. http://www.tpc.org/information/other/techarticles.asp.Google Scholar
- SPEC POWER. http://www.spec.org/specpower/.Google Scholar
- SPECCPU. http://www.spec.org/cpu2000/.Google Scholar
- SPECJBB. http://www.spec.org/jbb2005/.Google Scholar
- J. Stoess, C. Lang, and F. Bellosa. Energy Management for Hypervisor-Based Virtual Machines. In Proceedings of the 2007 USENIX Technical Conference (USENIX'07), June 2007. Google ScholarDigital Library
- B. Urgaonkar, P. Shenoy, and T. Roscoe. Resource Overbooking and Application Profiling in Shared Hosting Platforms. In Proceedings of the 5th USENIX Symposium on Operating Systems Design and Implementation (OSDI), Boston, Dec 2002. Google ScholarDigital Library
- C. Waldspurger. Memory Resource Management in VMWare ESX Server. In Proceedings of the Fifth Symposium on Operating System Design and Implementation (OSDI'02), December 2002. Google ScholarDigital Library
- X. Wang and M. Chen. Cluster-level feedback power control for performance optimization. In Proceedings of the Fourteenth International Symposium on High-Performance Computer Architecture (HPCA'08), February 2008.Google ScholarCross Ref
- Andreas Weisel and Frank Bellosa. Process cruise control-event-driven clock scaling for dynamic power management. In Proceedings of the International Conference on Compilers, Architecture and Synthesis for Embedded Systems, October 2002. Google ScholarDigital Library
- H. Zeng, X. Fan, C. Ellis, A. Lebeck, and A. Vahdat. ECOSystem: Managing Energy as a First Class Operating System Resource. In Proceedings of the Architectural Support for Programming Languages and Operating Systems, Oct 2002. Google ScholarDigital Library
- Server and Data Center Energy Efficiency -- EPA, August 2007. http://www.energystar.gov/ia/partners/prod development/downloads/EPA Datacenter Report Congress Final1.pdf.Google Scholar
Index Terms
- Statistical profiling-based techniques for effective power provisioning in data centers
Recommendations
How much power oversubscription is safe and allowed in data centers
ICAC '11: Proceedings of the 8th ACM international conference on Autonomic computingData centers attempt to maximize return on investment by achieving high levels of utilization. This means deploying the maximum number of servers possible within existing power supply capabilities. Therefore, a key problem is determining how many ...
Towards Economical Live Migration in Data Centers
Economics of Grids, Clouds, Systems, and ServicesAbstractLive migration of virtual machines (VMs) enables maintenance, load balancing, and power management in data centers. The cost of live migration on several key metrics combined with strict service-level objectives (SLOs), however, typically limits ...
A survey on virtual machine migration and server consolidation frameworks for cloud data centers
Modern Cloud Data Centers exploit virtualization for efficient resource management to reduce cloud computational cost and energy budget. Virtualization empowered by virtual machine (VM) migration meets the ever increasing demands of dynamic workload by ...
Comments