Dynamic Right-Sizing for Power-Proportional Data Centers
Top Cited Papers
- 22 November 2012
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE/ACM Transactions on Networking
- Vol. 21 (5), 1378-1391
- https://doi.org/10.1109/tnet.2012.2226216
Abstract
Power consumption imposes a significant cost for data centers implementing cloud services, yet much of that power is used to maintain excess service capacity during periods of low load. This paper investigates how much can be saved by dynamically “right-sizing” the data center by turning off servers during such periods and how to achieve that saving via an online algorithm. We propose a very general model and prove that the optimal offline algorithm for dynamic right-sizing has a simple structure when viewed in reverse time, and this structure is exploited to develop a new “lazy” online algorithm, which is proven to be 3-competitive. We validate the algorithm using traces from two real data-center workloads and show that significant cost savings are possible. Additionally, we contrast this new algorithm with the more traditional approach of receding horizon control.Keywords
This publication has 28 references indexed in Scilit:
- Optimality analysis of energy-performance trade-off for server farm managementPerformance Evaluation, 2010
- Optimality, fairness, and robustness in speed scaling designsPublished by Association for Computing Machinery (ACM) ,2010
- Energy-efficient algorithmsCommunications of the ACM, 2010
- Dynamic resource allocation and power management in virtualized data centersPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2010
- Optimal power allocation in server farmsPublished by Association for Computing Machinery (ACM) ,2009
- Multi-mode energy management for multi-tier server clustersPublished by Association for Computing Machinery (ACM) ,2008
- Optimal Power-Down StrategiesSIAM Journal on Computing, 2008
- Managing server energy and operational costs in hosting centersPublished by Association for Computing Machinery (ACM) ,2005
- Managing energy and server resources in hosting centersPublished by Association for Computing Machinery (ACM) ,2001
- Competitive snoopy cachingAlgorithmica, 1988