Spreadsheet Modeling of Data Center Hotspots

Spreadsheet Modeling of Data Center Hotspots

E.T.T. Wong (Department of Mechanical Engineering, The Hong Kong Polytechnic University, Hong Kong), M.C. Chan (Faculty of Business, The Hong Kong Polytechnic University, Hong Kong) and L.K.W. Sze (Department of Chinese and Bilingual Studies, The Hong Kong Polytechnic University, Hong Kong)
Copyright: © 2015 |Pages: 13
DOI: 10.4018/978-1-4666-5888-2.ch115
OnDemand PDF Download:
$30.00
List Price: $37.50

Chapter Preview

Top

Introduction

Today's data centers are evolving rapidly to embrace new technologies like the cloud, virtualization, consumer devices and more. As a result, data centers consume more energy than ever. A common management practice is that every data center periodically replaces its servers and major hardware during its technology refresh cycle. This permits organizations to apply updated technology to improve equipment reliability and enable new capabilities. Traditional data center refresh cycles took place about every five years, but owing to advanced technology, new guidelines for data processing environments, and alternative cooling technologies, they have been accelerating over the last decade.

In the case of advanced technology, a recent customer survey (Forrsights Hardware Survey, Q3 2011) revealed that among 161 enterprise IT infrastructure decision-makers across the US, the UK, and China, 57% have already adopted blade servers. Of those that have blade servers, nearly two-thirds are expanding their implementations (Forrester Consulting, 2012). Most users of blade servers reported increased server density and reduced power and cooling as the main operational and financial benefits. It can be anticipated that rapid adoption of refined blade technology will be forthcoming.

With regard to data processing environments, recently the American Society for Heating, Refrigerating and Air-Conditioning Engineers (ASHRAE) Technical Committee 9.9 described two new classes of data center equipment for which the maximum allowable temperature range is now 41 to 113°F (5 to 45ºC). This is a huge increase over ASHRAE’s 2008 recommended range, which listed the maximum allowable temperatures at 59 to 89.6°F. The move to higher operating temperatures means that there is a significant reduction in cooling needs. Hence more and more operators would prefer to have the enhanced equipment in place as soon as possible.

However, many data centers are not expanding physically at the same rate as their computing power. As a result the load density per square foot is increasing in virtually all data centers and this will continue into the foreseeable future. Many data centers are now in the 20-40 watts per square foot with some already approaching 100 watts per square foot. The increased heat load and heat density takes place in the form of more powerful servers located in the same rack configuration, and in the same physical space as before. This is because data center operators and owners cannot afford to relocate on a regular basis as property prices are increasing. The emerging problem is how to manage the increased heat density within the constraints of the existing data center and a promising approach to ensure the achievement of various availability tier targets is to model data center hardware defects and failures.

For most data centers, huge centralized air conditioning units that push air through drop ceilings or raised floors remain a regular phenomenon, but for enterprises building out for energy efficiency or seeking to retrofit for added energy relief, localized cooling—mainly in the form of in-row cooling systems—is making a start.

In-row cooling systems are finding their place between racks, pumping out cold air through the front and pulling in hot air from the back. Because cooling is performed by units just inches away from the heat source rather than indiscriminately through the floor or ceiling, data center hotspots run less hot. What is more, rather than relying on a central thermostat, these units function autonomously, tapping temperature-monitoring leads placed directly in front of a heat source to ensure that the air remains within a specified temperature range. Moreover, the unit ratchets down its cooling activities during partial loads. Nowadays, the cost-cutting benefits of localized cooling are quickly proving convincing, so much so that Gartner predicts in-rack and in-row cooling will become the predominant cooling method for the data center throughout the world.

However, as heat dissipation in data centers rises by orders of magnitude, inefficiencies such as air recirculation causing hotspots and flow short-circuiting has a significant impact on the system reliability and energy efficiency of the data center. Therefore, location and temperature distribution of hotspots is a significant challenge for data center management.

Key Terms in this Chapter

Blade Server: A server architecture that houses multiple server modules (“blades”) in a single chassis. It is widely used in data centers to save space and improve system management.

Availability: The ratio of the total time a data center system, subsystem or its components is capable of being used during a given interval (i.e. uptime) to the total duration of the interval (i.e. uptime + downtime).

Blade Technology: An integrated system designed to deliver server performance in efficient, high density, easy to expand and easy to manage units. It increases server density and facilitates server expansion. It also lowers management cost as it lowers cooling costs and reduces power consumption.

Uptime: A computer industry term for the time during which a computer is operational.

Downtime: A computer industry term for the time during which a computer is not operational.

Infrared Thermography: The process of acquisition and analysis of thermal information from non-contact thermal imaging devices.

Data Center (or Computer Center): A facility used to house computer systems and associated equipment, such as telecommunications and storage systems.

Server Refresh Cycle: The length of time that normally passes between installations of new servers and related hardware in a data center.

Reliability: The probability that a data center system or its components will perform its required function for a specified interval under stated conditions.

Bypass Airflow: Conditioned air that is short-cycled or not getting directly into the computer equipment within a data center.

Complete Chapter List

Search this Book:
Reset