Essential Rack
System Requirements
for Next Generation
Data Centers
White Paper #7
Revision 1
2003 American Power Conversion. All rights reserved. No part of this publication may be used, reproduced, photocopied, transmitted, or
stored in any retrieval system of any nature, without the written permission of the copyright owner.
www.apc.com
Rev 20031
2
Executive Summary
Effective mission critical installations must address the known problems and challenges
relating to current and past data center designs. This paper presents a categorized and
prioritized collection of rack system challenges and requirements as obtained through
systematic user interviews.
2003 American Power Conversion. All rights reserved. No part of this publication may be used, reproduced, photocopied, transmitted, or
stored in any retrieval system of any nature, without the written permission of the copyright owner.
www.apc.com
Rev 20031
3
Introduction
In the late 70’s companies began installing computers in rack enclosures; however, without multivendor
compatibility installations were difficult. In 1984 the International Electrotechnical Commission (IEC)
approved the IEC 2973 standard as a means to standardize the mechanical dimensions of 19inch (482.6
mm) enclosures. Other standards such as the EIA 310 (Electronic Industries Alliance) provided further
standardization to rack mounting telecommunications and IT equipment. Despite these standards, the
dynamic pace of the computer industry combined with its convergence with the telecom industry has caused
problems that these standards are slow to address. In this paper, a systematic approach of identifying and
classifying user problems provides insight regarding the nature and characteristics of rack systems in next
generation mission critical installations.
This paper focuses on the problems associated with rack mounting telecommunications equipment and IT
equipment in today’s data centers. A related APC white paper #4: “Essential Power System Requirements
for Next Generation Data Centers” addresses the related problems of providing power. APC white paper #5:
“Essential Cooling System Requirements for Next Generation Data Centers” focuses on the problem of
removing power in the form of heat.
Survey
A survey of management personnel relating to mission critical installations was conducted, interviewing
corporate CIO’s, Facility Managers, and IT Managers. Over 150 people were interviewed from over 90
different organizations including Fortune 1000 companies, Government and Education, and Service
Providers. Approximately 50% of customers interviewed were from North America, 20% from Europe, and
30% from Japan, Pacific, Australia, and Asia (JPAA) region.
The one year survey utilized “Voice of the Customer” techniques, which relies on data collection of verbal
and/or written responses to openended questions. This provides extremely unstructured responses, with
the advantage that the responses are not limited or constricted by preconceptions within the question.
During the course of the survey, some of the questions were expanded and/or changed in order to clarify
ambiguous responses.
Results: Rack System Challenges in Mission Critical
Installations
Survey responses were grouped according to common concepts, and for each group a solution requirement,
corresponding to a challenge for mission critical installation design, was derived. This process identified 18
core challenges. These core challenges were then further grouped according to theme into the following 5
key theme areas:
2003 American Power Conversion. All rights reserved. No part of this publication may be used, reproduced, photocopied, transmitted, or
stored in any retrieval system of any nature, without the written permission of the copyright owner.
www.apc.com
Rev 20031
4
•
Lifecycle Costs
•
Availability
•
Maintenance / Serviceability
•
Manageability
•
Adaptability / Scalability
For each theme area, the challenge, underlying problem, and power system requirements are presented in
tabular form. The highest priority problems are listed first under each theme. The priority was determined
by combining number of mentions weighted by priority as expressed by the respondents.
Lifecycle Cost Challenges
Challenge
Underlying problems
Rack System Requirements
Vendorneutral racks
Proprietary racks result in higher costs
due to “unique” requirements imposed
by IT equipment manufactures.
Difficulty with mergers and integrations
Vendorneutral rack manufacturer with economies of
scale and guaranteed universal compatibility with all
IT equipment.
Preengineered solutions that eliminate and/or
simplify most planning and engineering.
Accelerate speed of
deployment
The time and work involved in server
migrations and technology refreshes
are costly both in downtime and labor.
Preengineered time saving solutions that eliminate
and/or simplify planning and installation.
The survey found the lifecycle cost challenges were the most important requirement, particularly for
respondents from top level management in their organizations.
The survey revealed that most racks sold by server manufacturers are less likely to be compatible with
competitor’s equipment. This is a problem given that most respondents preferred to standardize on one rack
for a uniform and consistent look in the data center or network room as well as to eliminate the learning
curve of a different rack.
2003 American Power Conversion. All rights reserved. No part of this publication may be used, reproduced, photocopied, transmitted, or
stored in any retrieval system of any nature, without the written permission of the copyright owner.
www.apc.com
Rev 20031
5
Availability Challenges
Challenge
Underlying problems
Rack System Requirements
Proper cooling
airflow to IT
equipment
IT equipment compaction and rising heat
densities are causing equipment to fail
prematurely.
Servers are getting deeper and thinner
which allows more servers to be installed in
a rack and therefore more heat and cables.
Customers don’t know if the rack they are
using is providing proper airflow for their
equipment.
The temperature up and down the front of a
particular rack can vary 10
°
C (18
°
F). This
effect is unexpected and the reasons why
this happens are unclear to the users. This
places unexpected stress on individual
pieces of IT equipment and results in
premature failure of equipment above the
temperature gradient.
Welldesigned perforated doors should have
over 830 in
2
(0.53548 m
2
) of ventilation area
that can provide sufficient “unassisted” airflow
to ventilate IT equipment in a 42U rack.
Cable management should prevent power and
data cables from obstructing exhaust airflow.
Vendors that validate the cooling effectiveness
of their rack designs using CFD (Computational
Fluid Dynamics) and environmental chambers.
Tightly sealed rack enclosures and accessories
that prevent hot exhaust air from returning to
areas on the front of the rack, and assure that
cool supply air is distributed uniformly up and
down racks.
Provide dual power
sources to the
equipment
IT equipment today is available with single
and redundant power supplies but no
provision is made for bringing power
redundancy to the rack.
Easily configure the rack to provide dual power
paths to single or dual corded IT equipment.
Physical security
In trying to provide ample air, power and
data requirements, rack enclosures leave
critical equipment vulnerable to sabotage or
human error.
Hinges and fasteners located on the inside of
the doors to prevent access to equipment.
Doors should have unique combination locks
combined with a master key or electronic locks
integrated with the building security system.
Side panels should be keyed.
Seismic capability
(UBC)
Racks systems, located in Zone 4 regions
of the U.S., that are not designed to be in
compliance with the Universal Building
Code (UBC) for Zone4 seismic regions risk
catastrophic lose of the IT equipment they
are protecting.
All racks located in a Zone4 region should be
in compliance with UBC.
Minimize human
error
Shallow and constricted racks have a
propensity to exacerbate human error.
Rack based power strips that use small
unreliable circuit breakers can cause load
drops and represent another single point of
failure.
Resetting a circuit breaker inside a critical
rack can lead to further downtime.
Deep racks that allow ample room for working
with less fatigue and visual obstruction.
Remove circuit breakers from the IT rack.
Use rack based power strips without circuit
breakers for improved reliability.
Circuit breakers should be easily accessible
and located at the panel to avoid entering the
critical IT rack.
2003 American Power Conversion. All rights reserved. No part of this publication may be used, reproduced, photocopied, transmitted, or
stored in any retrieval system of any nature, without the written permission of the copyright owner.
www.apc.com
Rev 20031
6
The top availability problem for survey respondents was airflow to IT equipment. Given the dramatic
increase in heat densities over the last few years, IT administrators are concerned that IT equipment will
suffer damage due to inadequate airflow. To make matters worse, there exists no standard of measuring the
cooling effectiveness of one rack enclosure over another. This is needed to insure a highly available
environment for critical equipment. One method of ensuring proper cooling is to specify a rack doors that
provide over 830 in
2
(0.53548 m
2
) of ventilation area or doors that have a perforation pattern that is at least
63% open. Rack doors meeting these specifications can provide sufficient “unassisted” airflow to ventilate IT
equipment. For poor cooling environments, supplemental air moving accessories are recommended for
racks over 1,500 watts. The subject of rack cooling is addressed in detail in the following APC White
Papers:
White Paper 41: “Rack Cooling options for Data Centers and Network Rooms”
White Paper 44: “Improving Rack Cooling Performance Using Blanking Panels”
White Paper 40: “Heat trends within Information Technology Rack Enclosures”
Survey respondents had negative experiences with rack based power strips that used unreliable circuit
breakers. These inexpensive types of circuit breakers represent a single point of failure that is often
overlooked. Furthermore, if they trip, it forces human intervention inside a rack filled with critical equipment.
This risk is further increased by the fact that someone now has to locate the circuit breaker somewhere on
the power strip. These issues are mitigated by locating circuit breakers on the same panel thereby
increasing the availability of critical racks. By far the best way to increase power availability at the rack is to
bring power redundancy to the rack. Respondents were surprised by the dramatic availability increase when
dual power sources are brought to the rack, even for singlecorded IT equipment. This subject is discussed
in more detail in APC White Paper #48: “Comparing Availability of Various Rack Power Redundancy
Configurations”.
2003 American Power Conversion. All rights reserved. No part of this publication may be used, reproduced, photocopied, transmitted, or
stored in any retrieval system of any nature, without the written permission of the copyright owner.
www.apc.com
Rev 20031
7
Serviceability Challenges
Challenge
Underlying problems
Rack System Requirements
Decrease server
migration delays
Problems inevitably present themselves
which further delay completion of
migrations and technology refreshes.
Working clearances are restricted at
times thereby increasing the likelihood of
fatigue, injuries or downtime.
IT equipment installations can sometimes
take 3 to 4 times longer then normal due
to rack mounting difficulties.
Rack depths should allow for ample work room
thereby decreasing the likelihood of fatigue, injuries
and downtime.
Racks should incorporate toolless features such as
doors with quickrelease hinge pins, quickrelease
side panels, and power and cable management,
Racks should be able to roll through a standard
door.
Mounting rails should provide square holes instead
of round tapped holes.
Front and back mounting rails should have U
heights clearly marked to increase speed of
equipment deployment.
Racks should offer split rear doors, “French doors”,
to allow more working clearance in the aisles.
Cable
management
Power and data cables can obstruct
airflow causing damage to IT equipment
that isn’t properly cooled.
Difficulty in identifying power and data
cables due to “rats nest”.
Raised floors make it difficult to manage
power and data cables and usually
results in old cable being left under the
floor.
Racks should be designed with integrated channels
to allow for easy management, routing and storage
of large amounts of cable.
Data and power cables should be routed above the
rack for easy identification and serviceability.
Power and data cables should be stored in the rear
of the rack for easy access and management.
Standardized
racks for all
servers
Server manufactures often state that the
warrantee is void if a server is placed in a
rack other then their own. Apart from
aesthetics, this nonstandard approach
introduces complexity due to the unique
characteristics of each rack.
Racks should exceed the server manufacturers
ventilation and depth requirements.
Rack vendors should guarantee compatibility with all
servers.
Racks should comply with the EIA
310D standard.
Most of the serviceability challenges came from survey respondents that were directly involved with server
migrations and technology refreshes. It was evident that improvements in this area are highly dependant on
practical experience. Customers benefit tremendously from rack vendors who also provide cable
management and server migration services, since knowledge gained in the field is used to improve rack
designs.
2003 American Power Conversion. All rights reserved. No part of this publication may be used, reproduced, photocopied, transmitted, or
stored in any retrieval system of any nature, without the written permission of the copyright owner.
www.apc.com
Rev 20031
8
Manageability Challenges
Challenge
Underlying problems
Rack System
Requirements
Monitor environmental
variables at the rack
Difficulty in identifying thermal gradients
from the top to the bottom of the rack which
could shutdown and/or damage IT
equipment.
Difficulty in monitoring humidity at the rack.
Difficulty in detecting smoke at the very
early stages inside a rack.
Graphical user interfaces and
automatic notification which report,
manage, and notify based on
environmental parameters at the
rack level.
Monitor power attributes
at the rack level
Difficulty in determining racks that have high
thermal loads, and racks that are near
overload.
Difficulty in associating branch circuit loads
with racks due to constant reconfiguration.
For dual path systems, difficulty in
determining whether remaining circuits will
overload when one path goes down.
Graphical user interfaces and
automatic notification which report,
manage, and notify based on power
attributes at the rack level.
Ability to remotely and locally
monitor the current drawn from the
power strip(s) located inside each
rack. This is especially helpful after
adding new IT equipment.
Central management of IT
equipment
Expensive and difficult to individually
manage IT equipment that is distributed
more and more throughout facilities.
Software and hardware solutions
that allow IT administrators to
centrally manage all equipment.
Monitor security at the
rack
Racks are the last line of defense against
sabotage to IT equipment but are often
unmonitored.
Graphical user interfaces and
automatic notification which report,
manage, and notify security
breaches at the rack level.
Respondents focused mainly on environmental and power management inside the rack. These
management challenges closely resemble those discussed in APC white paper #4: “Essential Power System
Requirements for Next Generation Data Centers” and APC white paper #5: “Essential Cooling System
Requirements for Next Generation Data Centers”. However one unique management challenge that did
emerge from the surveys (central management) was based on the growing popularity of server clusters.
More companies are buying higher quantities of less expensive servers, which lower the cost of routine
operations and reduce single points of failure. According to IDC research vice president, Jean Bozman, the
developments in clustering technology have reduced the complexity involved in installing and maintaining
server farms. An IT administrator today can buy preconfigured clustered server systems that don’t require
specialized IT skills such as scripting. According to Dell senior manager of product marketing for clustering,
Sanjay Sidhu, clustering is making its way into missioncritical environments
1
. However, a KVM switch isn’t
enough to manage all these racks full of servers, IT personnel want a solution to centrally manage all
equipment from one location.
2003 American Power Conversion. All rights reserved. No part of this publication may be used, reproduced, photocopied, transmitted, or
stored in any retrieval system of any nature, without the written permission of the copyright owner.
www.apc.com
Rev 20031
9
Adaptability / Scalability Challenges
Challenge
Underlying problems
Rack System
Requirements
Plan for power density and
cooling requirements that are
increasing and unpredictable
Industry projections of equipment
power density and related cooling
requirements show great
uncertainty but new rack systems
must meet these requirements
despite equipment refreshes.
Rack design that can be easily
adapted, even retrofit, to house and
allow proper cooling for high density
equipment now and in the future.
Adapt to everchanging power
requirements
Different power requirements,
voltage requirements, outlet
requirements, even the need for
DC may occur at any time in any
rack.
A rack system that allows quick and
toolless changeover for different
voltages, power capacities, outlets,
and DC.
Allow for rear mounting of power strips
for easy power cord management.
Power strips should provide up to 42
receptacles.
Adapt to changing data center
and network room layouts
Sometime racks filled with
equipment must be moved to
locations that hamper the original
rack configuration with regard to
obstructions, and cabling.
Rack enclosures with field reversible
doors with quickrelease hinge pins,
quickrelease side panels, castors for
mobility, and scales up to 2,000 lb
(907.2 kg) capacity. Also, racks that
adapt to new overhead power and
data cabling systems.
Adapt to ever changing IT
equipment requirements
Telecom and Internet markets are
converging resulting in dynamic
enclosure requirements that leave
“holes” in racks. The cooling
environment of these rooms
become unstable.
A rack manufacture that can quickly
modify their racks and provide tool
less accessories to meet these
requirements.
Adjustable mounting rails.
The solution requirements to meet the Adaptability Challenges are based on preengineered and
standardized rack enclosures whose doors, side panels, rails, etc. can be easily changed in the field with no
tools. To satisfy the rack challenges identified in this survey, there are a number of changes required from
current design practice. Most of these changes arise by viewing the rack as the heart of the data center, for
it is at the rack level that a company’s critical IT processes are ultimately carried out. No longer should the
rack be thought of as just a metal enclosure and specified based on little more then price. The rack
enclosure provides a highly available environment that, when designed properly, can add measurable
probabilistic uptime and cost savings to a company. Indeed the rack should be considered part of the data
center or network room infrastructure which requires it be an integral part of power, cooling, security,
2003 American Power Conversion. All rights reserved. No part of this publication may be used, reproduced, photocopied, transmitted, or
stored in any retrieval system of any nature, without the written permission of the copyright owner.
www.apc.com
Rev 20031
10
management and fire detection systems. Only then will rack vendors have the foresight to provide solutions
to the problems identified by this survey as well as the problems yet to come.
Conclusions
A systematic analysis of customer problems relating to rack systems provides a clear statement of direction
for next generation mission critical installations. The most pressing problems that are not solved by current
design practices and equipment have the common theme of the inability of the data center or network room
to adapt to change. Rack systems must be more adaptable to changing requirements, in order to improve
both availability and cost effectiveness.
In many industries, a maturity level is reached where new advances in reliability, cycle time, and cost require
standardization, preengineering, and modularization. Designers of mission critical installations, designers of
the rack systems used in them, owners should consider whether this point has been reached. IDC predicts
that by 2005, 62% of all worldwide server unit shipments will be in a rackoptimized form factor. This puts
further pressure on the notion that the rack should be considered and designed as part of the mission critical
installation as the results of the survey in this paper suggest.
References
1. Lyman, J.,
Supporting Server Clusters
, News Factor Network, November 1, 2002.
www.newsfactor.com/perl/story/19843.html
Back to top
