IT systems have
for the day-to-day
operations of today’s
IT systems are critical for the day-to-day operations of
today’s organizations. Typically, IT systems within an
organization are housed at a centralized location often
referred to as the “server room” or the “data center.”
With many organizations experiencing continued server
sprawl or a significant increase in virtual machines, there
are increased pressures on the data center in terms of
cooling, electrical, airflow, and capacity. Whether you’re
building a new data center, evaluating your existing
internal data center, or evaluating a hosted (commonly
referred to as “colo”) data center, being proactive in
assessing the environment and capacity of the data
center can have a significant impact on the uptime,
thus impacting efficiency and effectiveness of an organization’s IT operations and thus, the overall business.
The availability and reliability of critical IT systems can be
largely dependent upon the data center’s design and
its operational and maintenance practices. Typically,
the availability of a data center is measured as a
percentage of uptime. Generally, uptime of 90 percent
or higher is viewed as being highly available; in the
case of IT system or data center uptime, however, anything short of 99 percent availability due to unplanned
interruptions is considered unacceptable (see Table 1).
These inadequacies prevented organizations from
responding to outages in a timely manner. According
to this study, the per-incident cost of a data center
outage can range between $38,969 and $1,017,746
per organization depending upon the industry and
size. (Industries surveyed ranged from public sector to
financial industries, and the costs included detection,
recovery, equipment, IT and end-user productivity,
third parties, lost revenue, and business disruption.)
In a separate survey conducted by the Aperture
Research Institute, the primary causes for data
center outages were: human error (57.3 percent),
improper controls/configurations (43.7 percent),
cooling issues (22.3 percent), and electrical issues
(26.2 percent). When coupled with best-in-class
operational and maintenance practices, proper
data center planning and design can help minimize
downtime at the data center.
Standardization in the data center
According to Wikipedia, a data center is defined as
“a facility used to house computer systems and
associated components, such as telecommunications
and storage systems. It generally includes redundant
Uptime % Est. Downtime/month Est. Downtime/year
1 day 19 hours 48 minutes
7 hours 12 minutes
3 days 15 hours 36 minutes
1 day 12 hours 0 minutes
18 days 6 hours 0 minutes
According to a study conducted by the Ponemon
Institute, approximately 95 percent of the surveyed
organizations experienced an unplanned data center
outage that was either attributed to inadequate capacity planning, design, or operational practices.
or backup power supplies, redundant data communications connections, environmental controls (e.g., air
conditioning, fire suppression), and security devices.”
If you looked up any definition of a data center, it
would typically contain terms such as redundancy,
capacity, power, cooling, monitoring, controls, and
security. So what makes one data center different from
another? It is the levels of redundancy, fault tolerance,
and scalability provided for cooling, electrical, connectivity, the capacity for space, operational procedures in
place, and other factors. In the absence of standards, it
would be hard to compare one data center to another.
This article focuses on the TIA-942 standard that can
be used when planning or comparing data centers.
This standard can be used as the benchmark and
may be applied to a new environment, retrofitting or
assessing an existing data center, or when evaluating
Standards provide a methodology for comparing
the capabilities of one data center to another and
avoiding the dangers of ambiguities or unsubstantiated qualitative references. The Uptime Institute (TUI)
provided a first attempt for standardization by
using a tiering system — a system that specifies the
availability, reliability, and capacity of a data center.
As the need for data center standardization gained
more traction, variations of these standards have
arisen. Currently, there are three commonly known tier
systems: (1) Uptime Institute TUI, (2) Syska Hennessy
Group, and (3) the TIA-942. A fourth standard by
ANSI/BICSI is also gaining some traction in this area.
The TIA-942 is a data center standard developed by
the Telecommunications Industry Association (TIA); the
TIA is accredited by the American National Standards
Institute (ANSI) to develop voluntary standards for
the telecommunications industry. Also referred to as
the ANSI/TIA-942, the TIA-942 standard’s purpose
is to provide design standards for a data center. This
standard takes multidisciplinary design approaches
and addresses 20 different areas, thereby promoting
cooperation between all parties involved in the design
and construction of the data center. This standard may
be used to benchmark an existing internal facility or a
The TIA-942 and TUI are the more commonly referenced standards; both provide a tier classification
system to help assess a site’s infrastructure. TUI and
Syska Hennessy Group provide a framework for the
disciplines in a data center but lack details, whereas
the TIA-942 provides some depth into various aspects
that go into a data center.
Given TIA-942 standard’s roots, this standard has a
strong emphasis on cabling specifications; however, it
also details requirements for architectural, electrical,
mechanical, security, and other site infrastructure
requirements. These 20 areas of the site infrastructure
can be broadly classified into the following seven core
areas, as shown below in Table 2.
What is the TIA-942?
Security & Protection
Access Provider & Demarcation
1. Design process
9. Fire protection
16. Access providers
2. Space planning
10. Physical security
17. Telecom space
3. Site selection
11. Building automation
18. Telecom administration
Architectural & Structural
2. Cabinets & racks
13. Cabling pathways
19. Information technology
(DRP, Ops Center)
izing & Selection
14. Cabling systems
Electrical & Mechanical
7. Electrical & grounding
8. HVAC/Mechanical systems
5. Cabling field testing
Based on the criticality or dependence upon the data
center, redundancy and fault tolerance need to be built
into the site’s infrastructure. The level of redundancy
and fault tolerance designed into each of these
20 areas of the data center’s infrastructure determines
its tier. The TIA-942 categorizes data centers into one
of four tiers (Tiers 1 – 4), where higher tiers correspond
to higher availability. Each individual area (mechanical,
electrical, cabling pathways, etc.) can be assessed
independently and categorized into a tier. For example, a data center may be rated Tier 4 for its electrical
characteristics but only Tier 3 for its mechanical characteristics. Just like the weakest link in a chain, a data
center’s overall tier rating is the lowest rating across all
aspects of its infrastructure. Annex G of the TIA-942
standard provides the tiering requirements for each of
these core areas.
What is the difference
between the TIA-942 and
The Uptime Institute standard?
The TIA-942 was developed by ANSI, an industry
association, whereas The Uptime Institute (TUI) is
a commercial for-profit entity. The TIA-942 tiering
scheme uses the TUI’s idea of a four-tiered system but
provides a prescriptive definition for each tier. The
TIA-942 classifies the tiers using quantifiable methods
while TUI uses a functional approach. TUI’s definition
of the tiering guideline is intentionally broad to allow
for innovation in achieving the desired level of a site’s
infrastructure performance. TUI tiers evaluate data
centers by their capability to allow maintenance and to
withstand a system or sub-system failure. Only the
TIA-942 offers a checklist form that can be used to
assess the tier of a data center.
The TIA-942 tiers are identified by the use of Arabic
numbers 1 through 4 (e.g., Tier 3), whereas the
Uptime Institute Tiers are identified by the use of
Roman numerals I through IV (e.g., Tier III). Both
standards identify recommendations for data center
design components, but the TIA-942 also covers
some operational and sustainability issues. The UTI
provides a second document (Tier Standard:
Operational Sustainability) to cover operational and
When it comes to the certification of a data center,
only the TUI offers a certification program. As of this
writing, TUI has awarded 149 certifications worldwide,
of which approximately 30 data centers are certified
as Tier IV centers based on their design documents,
constructed facility, or operational sustainability.
So, when data center vendors advertise that they’re a
Tier IV or Tier III certified data center, be aware how
rare that is. On the TIA-942 side, there’s no central
authority that certifies data centers based on the use
of the TIA-942 standard. Claims that a data center
is TIA-942 Tier 4 or Tier 3 (notice the use of Arabic
numbers) certified may be based on self-assessment
or a third-party audit of the TIA-942 checklist.
Both standards provide plenty of value, although
each has a slightly different philosophy. When
selecting a data center, it’s advisable to consider both
standards and then select the one that best meets
your business objectives.
What is a data center tier classification?
Both the TIA-942 and The Uptime Institute created the
tier classification to consistently describe the site-level
infrastructure required to sustain data center operations.
Each individual tier represents the requirements for the
data center design; the higher the tier, the greater the
sophistication of the design and operating requirements,
leading to increased data center availability.
The intent of this article is to give you a brief overview
of the TIA-942 data center standard. With this
understanding, you will have a basis to conduct further
research or to understand some of the claims made by
data center vendors. So, where do you go from here?
Assessing the criticality of the data center to your
organization will help identify the desired tier level of
your data center. If you have a low level of criticality,
does it make sense to invest in a Tier 3 or Tier 4 data
center? Probably not. Identifying the optimal criticality
for the organization identifies the required balance
between a data center’s cost of ownership and the
cost of downtime for the organization. Table 3 on the
next page provides the business characteristics that define criticality and the overall effect on system design.
Whatever the reason for your data center project (new
build, retrofit, or selection of a co-location data center),
it’s in your best interest to become familiar with the
terminology used in the data center industry to
understand any ambiguities that may come your way.
Any project involving the data center carries risk and
expense and therefore should not be taken lightly.
First, you need to perform a risk assessment to
determine hourly cost of downtime, which would
help you to identify the criticality of the data
center for your business.
Understand TIA-942 (requirement and process).
Assess the tier level of your current system and
Work with facilities to improve the overall health
of the data center.
Work with finance to justify upgrades. Alternatively, if it fits the business model, an outsourced
data center should be considered. Affordable
co-location or hosted data centers and high
uptime (99.995 percent) are not mutually exclusive.
Navigating these requirements can be daunting.
You may want to consider seeking assistance from
an outside professional before you act.
Judy Wright 248.223.3304
Sri Chalasani 248.223.3707
How can we help?
Whether you need help optimizing an existing IT infrastructure, implementing a new data center, or assessing a co-location facility to meet demands, Plante Moran can help. Professionals in the IT Infrastructure
group can assist you with:
Conducting a risk assessment and/or gap analysis of a current data center.
Providing data center design and capacity planning (infrastructure, electrical, HVAC, layout, etc.).
Working with architects in new/retrofit data center builds to ensure requirements are addressed.
Data center migration management.
Documenting an existing environment.
SSAE16 (SAS70) readiness preparation and audit management.
Recommended Data Center Tier
• Typically small business.
TIER 1 - Basic Infrastructure
• A small percentage of business revenue
• Non-redundant capacity components and distribution paths.
Numerous single points of failure in all aspects of design.
depends upon availability of IT systems. IT is
an enhancement to internal business processes.
• Business can function without IT system or
there is low dependence upon IT.
• View system downtime as a tolerable
• Principal use of web presence is for
• Any capacity component or distribution path failure will
impact the computer systems.
• Planned work will require most or all of the systems to
• Extremely vulnerable to inclement weather conditions.
• Generally unable to sustain more than a 10-minute
• There are no enforceable financial or non-financial
penalties due to loss of systems.
• Typically small to medium-sized business, mostly
limited to traditional 9-5 business hours.
• A small percentage of business revenue
depends upon availability of IT systems.
• Business relies upon availability of IT systems;
has multiple and often inter-related systems
• Some tolerance for downtime but extended
downtime becomes a business deterrent.
• Business does not provide real-time delivery
obligations or call centers with multiple sites.
TIER 2 - Redundant Components Infrastructure
• Redundant capacity components, but single non-redundant
distribution paths serving the site’s computer equipment
(i.e., a distribution path failure will cause the computer
equipment to shut down).
• Some redundancy in power and cooling systems.
• Redundant UPS modules and engine generators. Ability
to sustain a 24-hour power outage.
• Minimal thought in site selection.
• Formal data room separate from other areas.
• There are limited financial or non-financial
penalties due to loss of systems.
• Business serves both internal and external
• Business spans multiple time zones and
corresponding geographic diversity of
employees and customers.
• IT resources support automation of
• The customer impact due to short outages
• Business provides real-time delivery obligations
TIER 3 - Concurrently Maintainable Infrastructure
• Redundant capacity components and multiple distribution
paths serving the site’s computer equipment. Generally,
only one distribution path serves the computer equipment
at any time.
• Each and every capacity component and element of the
distribution paths can be removed from service on a planned
basis without causing any of the computer equipment to
be shut down.
• Ability to sustain a 72-hour power outage.
or call centers with multiple sites.
• Careful site selection planning process.
• Business has significant financial exposure
due to customer quality-of-service issues.
• Structure provides one-hour fire rating (how long it takes
flames to penetrate the structure).
• There are financial or non-financial penalties
due to loss of systems.
• Business operates in international markets,
with a “24 by forever” services commitment.
• Business processes and systems operation
• Majority of revenue comes from electronic
• Business needs to react to market transactions,
financial settlement, and e-commerce.
TIER 4 - Fault Tolerant Infrastructure
• Redundant capacity systems and multiple distribution paths
simultaneously serving the site’s computer equipment,
including utility power paths.
• There is 2N redundancy in power and cooling systems.
• Site is not susceptible to disruption of service from a single
unplanned worst-case event.
IT systems provides a competitive advantage.
• Infrastructure maintenance may be performed by using
redundant capacity components and distribution paths to
safely work on the remaining equipment.
• Extremely high cost of downtime.
• Ability to sustain a 96-hour power outage.
• Customer or employee access to applications/
• Stringent site selection criteria.
• Structure has a minimum two-hour fire rating.
• High level of physical security systems implemented.
• Availability of qualified staff to provide 24/7 onsite maintenance.