Abstract
In this paper, we propose UPAS (Utilization driven Power-Aware parallel job Scheduler) assuming DVFS enabled clusters. A CPU frequency assignment algorithm is integrated into the well established EASY backfilling job scheduling policy. Running a job at lower frequency results in a reduction in its power dissipation and energy consumption, but introduces a penalty in its performance. Furthermore, performance of other jobs may be affected as their wait times can increase. For this reason, we propose to apply DVFS when system utilization is below a certain threshold, exploiting periods of low system activity. As the increase in run times due to frequency scaling can be seen as an increase in computational load, we have done an analysis of HPC system dimension. This paper investigates whether having more DVFS enabled processors and scheduling jobs with UPAS can lead to lower energy consumption and higher performance. Five workload traces from systems in production use with up to 9 216 processors are simulated to evaluate the proposed algorithm and the dimensioning problem. Our approach decreases CPU energy by 8% on average depending on allowed job performance penalty. Applying UPAS to 20% larger systems, CPU energy needed to execute same workloads can be decreased by 20% while having same or better job performance.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
References
Butts J, Sohi G (2000) A static power model for architects. In: Proceedings of the 33rd annual IEEE/ACM international symposium on microarchitecture, MICRO-33, 2000
Dror Feitelson LR, Schwiegelshohn U (2005) Parallel job scheduling—A status report
Fan X, Weber W-D, Barroso LA (2007) Power provisioning for a warehouse-sized computer. In: ISCA ’07: Proceedings of the 34th annual international symposium on computer architecture
Feng X, Ge R, Cameron K (2005) Power and energy profiling of scientific applications on distributed systems. In: Proceedings of the 19th IEEE international symposium on parallel and distributed processing, 2005
Freeh V, Pan F, Kappiah N, Lowenthal D, Springer R (2005) Exploring the energy-time tradeoff in MPI programs on a power-scalable cluster. In: Proceedings of the 19th IEEE international symposium on parallel and distributed processing, 2005
Freeh VW, Lowenthal DK (2000) Using multiple energy gears in MPI programs on a power-scalable cluster. In: PPoPP ’05: Proceedings of the 10th ACM, SIGPLAN symposium on principles and practice of parallel programming
Freeh VW, Lowenthal DK, Pan F, Kappiah N, Springer R, Rountree BL, Femal ME (2007) Analyzing the energy-time trade-off in high-performance computing applications. IEEE Trans Parallel Distrib Syst
Guim F, Corbalan J (2008) A job self-scheduling policy for HPC infrastructures. In: JSSPPS ’08: Proceedings of the workshop on job scheduling strategies for parallel processing
Hikita J, Hirano A, Nakashima H (2008) Saving 200 kw and 200 k dollars per year by power-aware job and machine scheduling. In: IEEE international symposium on parallel and distributed processing, IPDPS 2008
Hotta Y, Sato M, Kimura H, Matsuoka S, Boku T, Takahashi D (2006) Profile-based optimization of power performance by using dynamic voltage scaling on a PC cluster. In: IPDPS, 2006
Hsu CH, Feng WC (2005) A power-aware run-time system for high-performance computing. In: SC, 2005
http://www.cs.huji.ac.il/labs/parallel/workload/. Parallel workload archive
Jackson D, Snell Q, Clement M (2001) Core algorithms of the Maui scheduler. In: Proceedings of the workshop on job scheduling strategies for parallel processing, 2001
Jones JP, Nitzberg B (1999) Scheduling for parallel supercomputing: a historical perspective of achievable utilization. In: IPPS/SPDP ’99/JSSPP ’99: Proceedings of the job scheduling strategies for parallel processing
Kamil S, Shalf J, Strohmaier E (2008) Power efficiency in high performance computing. In: IEEE international symposium on parallel and distributed processing, IPDPS 2008
Kappiah N, Freeh VW, Lowenthal DK (2005) Just in time dynamic voltage scaling: exploiting inter-node slack to save energy in MPI programs. In: SC, 2005
Kim KH, Buyya R, Kim J (2007) Power aware scheduling of bag-of-tasks applications with deadline constraints on DVS-enabled clusters. In: CCGRID, 2007
Lawson B, Smirni E (2005) Power-aware resource allocation in high-end systems via online simulation. In: ICS ’05: Proceedings of the 19th annual international conference on supercomputing
Lim MY, Freeh VW, Lowenthal DK (2006) Adaptive, transparent frequency and voltage scaling of communication phases in MPI programs. In: SC, 2006
Meisner D, Gold BT, Wenisch TF (2009) Powernap: eliminating server idle power. In: SIGPLAN Not, 2009
(Mootaz) Elnozahy EN, Kistler M, Rajamony R (2002) Energy-efficient server clusters. In: Proceedings of the 2nd workshop on power-aware computing systems, 2002
Mu’alem AW, Feitelson DG (2001) Utilization, predictability, workloads, and user runtime estimates in scheduling the IBM SP2 with backfilling. IEEE Trans Parallel Distrib Syst
Pinheiro E, Bianchini R, Carrera EV, Heath T (2001) Load balancing and unbalancing for power and performance in cluster-based systems. In: Workshop on compilers and operating systems for low power
Rountree B, Lowenthal DK, Funk S, Freeh VW, de Supinski BR, Schulz M (2007) Bounding energy consumption in large-scale MPI programs. In: SC ’07: Proceedings of the 2007 ACM/IEEE conference on supercomputing
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Etinski, M., Corbalan, J., Labarta, J. et al. Utilization driven power-aware parallel job scheduling. Comput Sci Res Dev 25, 207–216 (2010). https://doi.org/10.1007/s00450-010-0129-x
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00450-010-0129-x