We didn't start the fire: using an agent-directed thermal simulator to keep servers cool

ABSTRACT

As energy use by datacenters has risen over the years, the costs required to run a datacenter have substantially increased. Several thermal-aware algorithms exist to minimize energy consumption, but comparing these implementations can be difficult. Thermal modelers aid in the juxtaposition of several of these thermal-aware algorithms. Unfortunately, existing thermal modelers can be slow and difficult to use. Agent-Directed Thermal Modeler (ADTM) provides a solution that has a low learning curve and still quickly produces visualizations of a datacenter. The low number of input parameters significantly simplifies the use of ADTM, and its simulation time runs in seconds and minutes rather than hours. The resulting graphical and pictorial output can be used to determine which thermal-aware algorithm works best for a given datacenter.

ADTM is used to compare XInt-GA to random job placement in order to show up to a 13.6% increase in energy savings in an overloaded datacenter. This simulation for both algorithms takes only a few seconds, demonstrating that many thermal-aware algorithms can be compared quickly in order to determine the most effective and realistic algorithm that can be chosen to cool a datacenter.

References

1. Barjis, Joseph, Irina Rychkova, and Levent Yilmaz. 2011. "Modeling and simulation driven software development." In Proceedings of the 2011 Emerging M&S Applications in Industry and Academia Symposium, 4--10. Society for Computer Simulation International.
2. Brown, Richard, et al. 2008. "Report to Congress on server and data center energy efficiency: Public law 109--431."
3. Joonho, Kong, Chung Sung Woo, and Kevin Skadron. 2012. "Recent Thermal Management Techniques for Microprocessors." ACM Computing Surveys 44 (3): 13--13:42. issn: 03600300.
4. Kaushik, Rini T., and Klara Nahrstedt. 2012. "T*: a data-centric cooling energy costs reduction approach for big data analytics cloud." In Proceedings of the International Conference on High Performance Computing, Networking, Storage and Analysis, 52:1--52:11. SC '12. Salt Lake City, Utah: IEEE Computer Society Press. isbn: 978-1-4673-0804-5. http://dl.acm.org/citation.cfm?id=2388996.2389067.
5. Kennedy, James, and R Mendes. 2003. "Neighborhood topologies in fully-informed and best-of-neighborhood particle swarms." In Soft Computing in Industrial Applications, 2003. SMCia/03. Proceedings of the 2003 IEEE International Workshop on, 45--50. IEEE.
6. Moss, David. 2009. "Data Center Operating Temperature: What Does Dell Recommend?" http://www.dell.com/Learn/us/en/19/business~solutions~whitepapers~en/Documents~dci-Data-Center-Operating-Temperature-Dell-Recommendation.pdf?c=us&l=en&s=dhs.
7. Mukherjee, Tridib, Ayan Banerjee, Georgios Varsamopoulos, Sandeep K. S. Gupta, and Sanjay Rungta. 2009. "Spatio-temporal thermal-aware job scheduling to minimize energy consumption in virtualized heterogeneous data centers." Computer Networks 53 (17): 2888--2904. issn: 13891286.
8. Tang, Qinghui, Sandeep Kumar S Gupta, and Georgios Varsamopoulos. 2008. "Energy-efficient thermal-aware task scheduling for homogeneous high-performance computing data centers: A cyber-physical approach." Parallel and Distributed Systems, IEEE Transactions on 19 (11): 1458--1472.
9. Team, Repast Development. 2012. "Repast Suite." http://repast.sourceforge.net/repast_simphony.html.

Index Terms

• Computing methodologies

  • Artificial intelligence

    • Distributed artificial intelligence

      • Multi-agent systems

  • Modeling and simulation

    • Simulation evaluation

    • Simulation types and techniques