A quantitative analysis of cooling power in container-based data centers

Cooling power is often represented as a single taxed cost on the total energy consumption of the data center. Some estimates go as far as 50% of the total energy demand. However, this view is rather simplistic in the presence of a multitude of cooling options and optimizations. In response to the rising cost of energy, the industry introduced modular design in the form of containers to serve as the new building block for data centers. However, it is still unclear how efficient they are compared to raised-floor data centers and under what conditions they are preferred. In this paper, we provide comparative and quantitative analysis of cooling power in both container-based and raised-floor data centers. Our results show that a container achieves 80% and 42% savings in cooling and facility powers respectively compared to a raised-floor data center and that savings of 41% in cooling power are possible when workloads are consolidated onto the least number of containers. We also show that cooling optimizations are not very effective at high utilizations; and that a raised-floor data center can approach the efficiency of a container at low utilizations when employing a simple cooling optimization.

[1]  T. N. Vijaykumar,et al.  Joint optimization of idle and cooling power in data centers while maintaining response time , 2010, ASPLOS XV.

[2]  Kashi Venkatesh Vishwanath,et al.  Modular data centers: how to design them? , 2009, LSAP '09.

[3]  Arnold O. Allen,et al.  Probability, statistics and queueing theory - with computer science applications (2. ed.) , 1981, Int. CMG Conference.

[4]  Jeffrey S. Chase,et al.  Making Scheduling "Cool": Temperature-Aware Workload Placement in Data Centers , 2005, USENIX Annual Technical Conference, General Track.

[5]  Gunter Bolch,et al.  Queueing Networks and Markov Chains - Modeling and Performance Evaluation with Computer Science Applications, Second Edition , 1998 .

[6]  Luiz André Barroso,et al.  The Datacenter as a Computer: An Introduction to the Design of Warehouse-Scale Machines , 2009, The Datacenter as a Computer: An Introduction to the Design of Warehouse-Scale Machines.

[7]  Aarnout Brombacher,et al.  Probability... , 2009, Qual. Reliab. Eng. Int..

[8]  Arnold O. Allen,et al.  Probablity, Statistics and Queueing Theory with Computer Science Applications, Second Edition , 1990, Int. CMG Conference.

[9]  Thomas F. Wenisch,et al.  PowerNap: eliminating server idle power , 2009, ASPLOS.

[10]  M.K. Patterson,et al.  The effect of data center temperature on energy efficiency , 2008, 2008 11th Intersociety Conference on Thermal and Thermomechanical Phenomena in Electronic Systems.

[11]  Richard E. Brown,et al.  Report to Congress on Server and Data Center Energy Efficiency: Public Law 109-431 , 2008 .

[12]  Cullen E. Bash,et al.  Viability of Dynamic Cooling Control in a Data Center Environment , 2006 .

[13]  Faraz Ahmad,et al.  Joint optimization of idle and cooling power in data centers while maintaining response time , 2010, ASPLOS 2010.

[14]  James R. Hamilton,et al.  An Architecture for Modular Data Centers , 2006, CIDR.

[15]  Luiz André Barroso,et al.  The Case for Energy-Proportional Computing , 2007, Computer.

[16]  Christian Belady,et al.  GREEN GRID DATA CENTER POWER EFFICIENCY METRICS: PUE AND DCIE , 2008 .

[17]  Jeffrey S. Chase,et al.  Balance of power: dynamic thermal management for Internet data centers , 2005, IEEE Internet Computing.

[18]  R. Sharma,et al.  Experimental investigation of design and performance of data centers , 2004, The Ninth Intersociety Conference on Thermal and Thermomechanical Phenomena In Electronic Systems (IEEE Cat. No.04CH37543).

[19]  Myron Hlynka,et al.  Queueing Networks and Markov Chains (Modeling and Performance Evaluation With Computer Science Applications) , 2007, Technometrics.

[20]  Wolf-Dietrich Weber,et al.  Power provisioning for a warehouse-sized computer , 2007, ISCA '07.